Jurkat T-Cell Antigen-Independent Elimination of PMA-Activated Neuroblastoma Cells Is Triggered by CCL2/CCR2, Depends Upon Lipid Raft LFA1/ICAM1 Immune Synapses, Is Mediated by m-TRAIL and Is Augmented by the TrkAIII Oncoprotein

TL;DR

This study identifies a new immune mechanism for killing neuroblastoma cells using Jurkat T-cells, involving CCL2/CCR2 and immune synapses, which could lead to new immunotherapies for high-risk neuroblastomas.

Contribution

The study reveals a novel antigen-independent cytotoxic mechanism involving CCL2/CCR2 and lipid raft immune synapses for targeting neuroblastoma cells.

Findings

PMA-activated neuroblastoma cells trigger Jurkat T-cell cytotoxicity via CCL2/CCR2 signaling and LFA1/ICAM1 immune synapses.

The mechanism is enhanced by the TrkAIII oncoprotein and is effective against both MYCN-amplified and non-amplified neuroblastoma cells with PMA-inducible CCL2.

The process is resistant to osteoprotegerin and PD-L1/PD-1 but is offset by Fas-mediated Jurkat cell apoptosis.

Abstract

Advances in multimodal therapy for high-risk neuroblastomas (NBs) have plateaued, prompting therapeutic initiatives to harness the immune system. NBs, however, are immunologically “cold” and a significant challenge to immunotherapy. Here, in a Jurkat lymphocyte cytotoxicity model, we describe an antigen-independent, cell-mediated mechanism for eliminating NB cells, first detected in PMA-activated low pcDNA-SH-SY5Y and high TrkAIII-SH-SY5Y TrkAIII-expressing cells, which are resistant to Jurkat elimination under normal conditions. Characterization of this mechanism through live cell imaging, adhesion assays, RT-PCR, Western blotting and indirect IF, employing a variety of inhibitors, indicates that it initiates with PMA-induced NB cell CCL2 expression. This results in CCL2 promotion of Jurkat CCR2b expression, CCL2/CCR2b-mediated Jurkat LFA-1 activation and the formation of cytotoxic lipid-raft LFA1/ICAM-1 immune synapses, through which Jurkat m-TRAIL combines with PMA-enhanced NB cell DR5/TRAIL-R2 expression to induce NB cell apoptosis. This mechanism is enhanced by the NB-associated oncoprotein TrkAIII through Shp/Src-regulated c-FLIP sequester and is PD-L1/PD-1-independent and resistant to osteoprotegerin. It eliminates both non-MYCN-amplified (SH-SY5Y and SK-N-SH) and MYCN-amplified (SMS-KCNR) NB cells that exhibit PMA-inducible CCL2 expression but not MYCN-amplified NB cells (IMR-32 and NB-1) that exhibit CCL2 repression, and is offset by reciprocal NB cell-induced Fas-mediated Jurkat cell apoptosis. These findings form a solid foundation for further pre-clinical development aimed at identifying clinically relevant physiological immune cell equivalents and alternative PKC activators, with the ultimate goal of translating this mechanism into an effective immune-therapeutic approach for the treatment of high-risk non-immunogenic NBs, especially NBs that exhibit CCL2 and TrkAIII expression.

Figures

Full text

1. Introduction

Neuroblastomas (NBs) are heterogeneous, aggressive, therapy-resistant embryonal tumors of sympathoadrenal progenitor cell-committed, neural crest origin, represent the most prevalent extracranial solid tumor of childhood and account for approximately 15% of pediatric cancer-related deaths [1,2,3]. Advances in multimodal therapy for high-risk disease have increased 5-year survival rates to around 50% but have now reached a plateau, primarily due to issues of cytotoxicity and immune depression, prompting alternative and/or complementary initiatives aimed at harnessing the immune system. NBs, however, are immunologically “cold” tumors and represent a significant immunotherapeutic challenge [4,5,6,7]. Despite low antigenicity and immunogenicity, 28 to 100% of NBs exhibit infiltration by varying percentages of functional α/β and γ/δ T cells, NK and variant iNK cells [7,8,9,10], which have been reported to increase in association with response to chemotherapy, consistent with therapy-induced immunogenic death [11,12,13,14]. Furthermore, despite reports of an inverse correlation between MYCN amplification and infiltration by activated lymphocytes, MYCN expression in non-MYCN amplified NBs has been associated with elevated lymphocyte infiltration and better outcomes, highlighting the immunotherapeutic potential of promoting cytotoxic lymphocyte infiltration, particularly in non-MYCN amplified NB [15,16,17,18].

NB metastatic progression, high-risk classification and post-therapeutic relapse are associated not only with MYCN gene amplification, which is detected in 20–30% of all NBs and approximately 38% of aggressive, advanced-stage metastatic NBs [19,20], but also with the expression of the oncogenic alternative TrkAIII splice variant (GeneBank: OP866787.1) of the NGF neurotrophin receptor, tropomyosin-related kinase A (TrkA/NTRK1) [21,22]. Both MYCN amplification and alternative TrkAIII splicing robustly predict outcomes [21,22]. The TrkAIII variant mRNA exhibits in-frame skipping of cassette exons 6 and 7 that encode protein domains required for cell surface receptor expression, prevention of spontaneous activation and optimization of ligand-binding [21,23]. As a consequence, TrkAIII is not expressed at the cell surface but relocalizes to pre-Golgi membranes, centrosomes and mitochondria, where it exhibits ligand-independent cell-cycle and stress-regulated activation [21,23,24,25] and promotes primary and metastatic tumorigenicity in NB models [21,23,26]. TrkAIII expression in NB cells is promoted by tumor microenvironmental conditions, including hypoxia, nutrient deprivation and ER stress [21,23,27], and actionable potential is supported by its association with advanced-stage metastatic NB and post-therapeutic relapse [21,22], its equivalent oncogenic activity to the TrkT3 TrkA-fusion oncogene in NB models [21], and by a report of a dramatic durable response to the Trk inhibitor entrectinib in an infant with refractory stage IV metastatic NB, exhibiting molecular profile consistent with TrkAIII expression and activation [28].

Paradoxically, TrkAIII also sensitizes NB cells to apoptosis induced by soluble (s) TNF-related apoptosis-inducing ligand (s-TRAIL), unveiling a potential immunotherapeutic Achilles heel [29]. This effect depends upon TRAIL-induced, Shp and Src-regulated TrkAIII sequestration of the non-catalytic caspase-8 homologue c-FLIP, reducing the availability of c-FLIP to inhibit the death-inducing signaling complex (DISC) and prevent TRAIL receptor recruitment into lipid rafts, both of which are prerequisites for canonical TRAIL/TRAIL-R apoptotic signaling [30,31,32]. This suggests that NBs that express TrkAIII may respond to TRAIL therapy.

Therapeutic interest in TRAIL stems from its capacity to induce apoptosis in tumor cells but not in normal cells, and the high degree of toleration and safety reported in clinical trials [33,34,35]. Therapeutic TRAIL formulations, however, are hampered by the need to stabilize TRAIL in functional trimeric form [36], rapid s-TRAIL clearance and short half-life in circulation [37,38], problems of tumor delivery and penetration, and evasion mechanisms, including anti-apoptotic equilibria between functional TRAIL receptors, decoy TRAIL receptors and cFLIP, as well as the TRAIL inhibitor osteoprotegerin, particularly within the NB bone metastatic niche [39,40,41,42,43,44,45,46]. TRAIL, however, is also expressed in a significantly more potent, trimer-stabilized, membrane-bound form (m-TRAIL) [36,40,47], which has been reported to enhance NK elimination of NB cells [48] and to inhibit the metastatic state in progressive NBs [49]. This suggests that immune cells that express this more potent form of TRAIL may be more effective at eliminating TRAIL-sensitive NB cells, by improving TRAIL stability, tumor penetration and delivery.

Within this context, and based on our previous observations [29], we evaluated the sensitivity of low- and high-TrkAIII-expressing SH-SY5Y NB cells to cell-mediated cytotoxicity in a Jurkat T-cell model [50,51,52]. Although Jurkat CD4^+^ leukemia T cells resemble helper rather than cytotoxic CD8^+^ T cells, do not exhibit antigen-dependent cytotoxicity, express low levels of FasL and Granzyme B [50,51,52,53,54,55,56,57] and do not perfectly replicate cytotoxic lymphocyte behavior, they do express TRAIL (this study) [58], and induce tumor cell apoptosis under different conditions [48,52,53]. This makes this well-characterized cell line a valuable tool for evaluating the sensitivity of non-immunogenic NB cells to antigen-independent, cell-mediated cytotoxicity.

2. Results

2.1. PMA Reduces pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y Proliferation

Cell counts from Incucyte live cell imaging micrographs confirmed a doubling time for Jurkat T-cells of 19.6 ± 0.13 h, for untreated pcDNA-SH-SY5Y cells of 14.5 ± 0.15 h and for untreated TrkAIII-SH-SY5Y cells of 14.8 ± 0.09 h. PMA treatment (60 ng/mL for 16 h, followed by removal) significantly increased pcDNA-SH-SY5Y doubling time to 24.3 ± 0.5 h and TrkAIII-SH-SY5Y doubling time to 24.7 ± 0.73 h, in the absence of significant cell death.

2.2. PMA Sensitization of SH-SY5Y Cells to Elimination by Jurkat Cells Is Augmented by TrkAIII

Sub-confluent non-activated TrkAIII-SH-SY5Y and pcDNA SH-SY5Y cells were resistant to elimination by Jurkat cells, assessed over 48 h by live cell imaging at Jurkat to NB cell ratios of 1 to 2; 1 to 1; 2 to 1; and 4 to 1 (Figure 1a, displayed for TrkAIII-SH-SY5Y cells only at a 2 to 1 ratio).

PMA-activated pcDNA-SH-SY5Y cells (60 ng/mL for 16 h, followed by removal) were not significantly eliminated by Jurkat cells at 24 h, at ratios of 1 to 2 (0% elimination) or 1 to 1 (1.6 ± 0.5% elimination) with NB cells. They were partially eliminated at 24 h by Jurkat cells at a ratio with NB cells of 2 to 1 (50.8 ± 4.3% elimination) and almost completely eliminated at 24 h by Jurkat cells at a ratio with NB cells of 4 to 1 (83 ± 8.7% elimination) (Figure 1b). In contrast, PMA-activated TrkAIII-SH-SY5Y cells (60 ng/mL for 16 h, followed by removal) were almost completely eliminated at 24 h by Jurkat cells at a ratio with NB cells of 1 to 2 (78.9 ± 8.3% eliminated) and completely eliminated by Jurkat cells at ratios with NB cells of 1 to 1 (98.7 ± 5.7% elimination), 2 to 1 (98.9 ± 4.9% elimination), and 4 to 1 (98.8 ± 0.9% elimination) (p < 0.0001, for all ratios at 24 h) (Figure 1b,c).

Live cell imaging confirmed that Jurkat elimination of PMA-activated pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells was predominantly associated with Jurkat recruitment and adherence to NB cells. Furthermore, cell-free 24 h-conditioned medium from Jurkat co-cultures with PMA-activated TrkAIII-SH-SY5Y cells, exhibiting maximal TrkAIII-SH-SY5Y cell death, did not induce significant PMA-activated TrkAIII-SH-SY5Y cell death (Figure 1c).

These data indicate that Jurkat elimination of PMA-activated pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells is cell-mediated, TrkAIII expression augments the sensitivity of PMA-activated SH-SY5Y cells to elimination by Jurkat cells and PMA-activated pcDNA-SH-SY5Y cell resistance to Jurkat elimination is overcome by increasing the Jurkat to NB cell ratio.

2.3. Molecular Characterization of Jurkat Elimination of PMA-Activated SH-SY5Y Cells

2.3.1. Jurkat Elimination of PMA-Activated TrkAIII-SH-SY5Y Cells Is Mediated by m-TRAIL

Indirect IF confirmed that Jurkat cells express TRAIL, which was localized to intracellular vesicles in fixed and permeabilized Jurkat cells and was not detected at the cell surface in live, non-permeabilized Jurkat cells (Figure 2a).

In live cell imaging assays, the neutralizing anti-human TRAIL antibody [59] (10 μg/mL) significantly prevented Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells at a Jurkat to NB cell ratio of 2 to 1, by 93.5% at 6 h, 89.8% at 12 h and 87.5% at 24 h (p < 0.0001 for all 3 time points). In contrast, the peptide Fas inhibitor MET12 (10 μM) [60] failed to significantly reduce the elimination of PMA-activated TrkAIII-SH-SY5Y cells at 6, 12 and 24 h, at the same Jurkat to NB cell ratio (Figure 2b,c).

These data indicate that in the absence of PMA-activated NB cells, Jurkat TRAIL localizes to intracellular vesicles and that Jurkat elimination of PMA-activated SH-SY5Y cells is m-TRAIL-mediated and does not involve Fas.

2.3.2. Jurkat Elimination of PMA-Activated TrkAIII-SH-SY5Y Cells Occurs via the Extrinsic Apoptosis Pathway

Western blots detected caspase-8 and caspase-3 but not caspase-9 cleavage in whole cell extracts from 24 h co-cultures of Jurkat and PMA-activated TrkAIII-SH-SY5Y cells at a 2 to 1 ratio, but did not detect either caspase-3 or caspase-8 cleavage in whole cell extracts from 24 h co-cultures of Jurkat and non-activated TrkAIII-SH-SY5Y cells at the same ratio (Figure 3a).

In live cell imaging assays, the pan-caspase inhibitor z-VAD-fmk (10 μM) [61] reduced Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells by 98.8% at 6 h, 99.3% at 12 h and 96.7% at 24 h (p < 0.0001 for all time points). The caspase-8 inhibitor z-IETD-fmk (10 μM) [62] reduced Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells by 100% at 6 h, 99.3% at 12 h and 98.9% at 24 h (p < 0.0001 for all 3 time points), and the caspase-3 inhibitor z-DEVD-fmk (10 μM) [63] reduced Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells by 100% at 6 h, 92.2% at 12 h and 54.7% at 24 h (p < 0.0001 for all 3 time points). In contrast, Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells was not significantly reduced by the Htr2A/Omi inhibitor UCF-101 (1 μg/mL) [64] at either 6, 12 or 24 h (Figure 3b).

2.3.3. Jurkat Elimination of PMA-Activated TrkAIII-SH-SY5Y Cells Depends upon LFA-1 and ICAM-1

Western blotting and indirect IF detected constitutive CD11a and ICAM-1 protein expression and immunoreactivity in Jurkat cells, as well as constitutive ICAM-1 protein expression and immunoreactivity in TrkAIII-SH-SY5Y cells. Treatment with PMA (60 ng/mL for 16 h) did not increase TrkAIII-SH-SY5Y ICAM-1 protein expression (Figure 4a).

In adhesion assays, Jurkat cell adhesion to non-activated TrkAIII-SH-SY5Y cells was significantly increased by 1.91 (±0.12)-fold (p > 0.0001) to PMA-activated (60 ng/mL for 16 h followed by PMA removal) TrkAIII-SH-SY5Y cells. Pre-incubation of Jurkat cells with the LFA-1 inhibitor BIRT377 [65] (15 μM for 30 min, followed by removal) did not significantly reduce Jurkat adhesion to non-activated TrkAIII-SH-SY5Y cells but significantly reduced Jurkat adhesion to PMA-activated TrkAIII-SH-SY5Y cells (60 ng/mL for 16 h, followed by PMA removal) from 1.91 ± 0.12 to 0.83 ± 0.21, representing a ≈ 67% reduction (p < 0.0001) (Figure 4b).

In live cell imaging assays, BIRT377 (15 μM, Jurkat pre-incubation for 30 min and throughout the assay) reduced Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells, at a 2 to 1 ratio, by 86.2% at 6 h, 90.4% at 12 h and 54.3% at 24 h (p < 0.0001, for all 3 time points). Human recombinant ICAM-1 (10 μg/mL, Jurkat cell pre-incubation for 30 min and present throughout the assay) also reduced Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells by 98.8% at 6 h, 93.7% at 12 h and 75.8% at 24 h (p < 0.0001 for all 3 time points) (Figure 5a,b).

These data indicate that Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells depends upon LFA-1 and ICAM-1 and involves Jurkat LFA-1 activation and interaction with SH-SY5Y ICAM-1.

2.3.4. LFA-1 and ICAM-1 Involvement in Jurkat Elimination of PMA-Activated TrkAIII-SH-SY5Y Depends upon the CCL2/CCR2 Chemokine Axis

RT-PCR detected low-level constitutive CCL2 mRNA expression in non-activated TrkAIII-SH-SY5Y but not pcDNA-SH-SY5Y cells. PMA (60 ng/mL for 6 h) induced CCL2 mRNA expression in pcDNA-SH-SY5Y cells and enhanced CCL2 mRNA expression in TrkAIII-SH-SY5Y cells (Figure 6a). Western blotting confirmed induction of pcDNA-SH-SY5Y CCL2 protein levels and enhanced TrkAIII-SH-SY5Y CCL2 protein levels, relative to MMP2, in 20-fold concentrated 48 h serum-free conditioned medium collected following PMA treatment (60 ng/mL for 16 h, followed by removal) compared to untreated counterparts (Figure 6b).

RT-PCR also detected low-level constitutive CCR2b mRNA expression in Jurkat cells, similar CCR2b mRNA levels in Jurkat cells isolated following a 6 h interaction with non-activated TrkAIII-SH-SY5Y cells and notably enhanced CCR2b mRNA expression in Jurkat cells isolated following a 10 h interaction with PMA-activated TrkAIII-SH-SY5Y cells (60 ng/mL for 16 h, followed by removal) (Figure 6c). Western blots and indirect IF also detected increased CCR2 protein expression and immunoreactivity, respectively, in Jurkat cells isolated following a 10 h interaction with PMA-activated (60 ng/mL for 16 h, followed by removal) compared to non-activated TrkAIII-SH-SY5Y cells (Figure 6d). CCR2 mRNA expression was not detected in either non-activated or PMA-activated (60 ng/mL for 6 h) pcDNA-SH-SY5Y or TrkAIII-SH-SY5Y cells.

RT-PCR detected CXCR4 mRNA expression in Jurkat cells alone and in Jurkat cells isolated following a 10 h interaction with non-activated TrkAIII-SH-SY5Y cells. In contrast, in Jurkat cells isolated following a 10 h interaction with PMA-activated TrkAIII-SH-SY5Y cells (60 ng/mL for 16 h, followed by removal), CXCR4 mRNA expression was markedly reduced and was not restored in the presence of bindarit (100 μM) (Figure 7a).

Western blots and indirect IF also detected a marked reduction in CXCR4 protein expression and immunoreactivity in Jurkat cells isolated following a 10 h interaction with PMA-activated TrkAIII-SH-SY5Y cells (60 ng/mL for 16 h, followed by removal) compared to Jurkat cells isolated following a 10 h interaction with non-activated TrkAIII-SH-SY5Y cells (Figure 7b).

RT-PCR detected CXCL12 mRNA expression in non-activated pcDNA-SH-SY5Y cells, which was not enhanced by PMA treatment (60 ng/mL for 6 h) but did not detect CXCL12 mRNA expression in either non-activated or PMA-activated (60 ng/mL for 6 h) TrkAIII-SH-SY5Y cells (Figure 7c).

The CCL2 inhibitor bindarit [66] (100 μM) markedly reduced CCL2 levels, relative to MMP-2, in 20× concentrated 48 h serum-free conditioned media from PMA-activated TrkAIII-SH-SY5Y cells (60 ng/mL for 16 h followed by removal and 48 h serum-free culture) (Figure 7d). In live cell imaging assays, bindarit (100 μM, present during PMA activation of NB cells, then removed for assay) significantly inhibited Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells by ≥98% at 6, 12 and 24 h time points (p < 0.0001, for all 3 time points). The CCR2 inhibitor INCB3284 [67] (100 nM, Jurkat pre-incubation for 30 min, then present throughout the assay) also significantly reduced Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells by 100% at 6 h, 61.5% at 12 h and 23.2% at 24 h (p < 0.0001 for all 3 time points). In contrast, the CXCR4 inhibitor IT1t-dihydrochloride [68] (10 μM, Jurkat pre-incubation for 30 min and present throughout the assay) did not significantly reduce Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells (Figure 7e).

In order to exclude any potential influence of residual PMA on Jurkat cell activation, estimated to be ≤0.002 ng/mL according to our washing protocol, considering that Jurkat interaction with PMA-activated TrkAIII-SH-SY5Y cells induced Jurkat CCR2b expression, we evaluated the effect of 6 h treatment with PMA at 60 ng/mL and 0.002 ng/mL on Jurkat CCR2b mRNA levels. As demonstrated in Supplementary Figure S1, Jurkat CCR2b mRNA expression, barely detectable in untreated controls, was markedly induced by 60 ng/mL PMA but not by 0.002 ng/mL PMA. This excludes residual PMA involvement in Jurkat CCR2b expression induced upon co-culture with PMA-activated TrkAIII-SH-SY5Y cells.

These data confirm that the CCL2/CCR2 but not CXCL12/CXCR4 chemokine axis is involved in Jurkat LFA-1/ICAM-1-dependent elimination of PMA-activated TrkAIII-SH-SY5Y cells.

2.4. Jurkat Elimination of PMA-Activated NB Cells Is Associated with the Formation of LFA-1/ICAM-1 Immune Synapses

Indirect IF detected Jurkat CD11a and TrkAIII-SH-SY5Y ICAM-1 and DR5/TRAIL-R2 clustering with ring formation, SH-SY5Y γ-tubulin (centrosome) and GM153 (Golgi) relocalization close to ICAM-1 and DR5 clusters, and the polarization of Jurkat TRAIL vesicles towards interaction sites between Jurkat cells and PMA-activated TrkAIII-SH-SY5Y cells, but not between Jurkat cells and non-activated TrkAIII-SH-SY5Y cells, in 6 h co-cultures (Figure 8).

In live cell imaging assays, the lipid raft-disrupting agent methyl-β-cyclodextrin [69] (100 μM, NB cell pre-incubation for 30 min, then removed for the assay) significantly reduced the Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells, by 92.4% at 6 h, 66.1% at 12 h and 45.4% at 24 h, at a ratio of 2 to 1 (p < 0.0001 for all 3 time points) (Figure 9a,b).

These data implicate cytotoxic lipid-raft LFA-1/ICAM-1 immune synapses in the Jurkat elimination of PMA-activated TrkAIII SH-SY5Y cells.

2.5. PMA Promotes a Pro-Apoptotic Equilibrium Between TRAIL Receptors, c-FLIP and Caspase-8

Western blotting detected a significant increase in DR5 (TRAIL-R2) protein levels in whole cell extracts from PMA-activated cells (60 ng/mL for 16-h) compared to non-activated pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells but did not detect significant changes in DR4 (TRAIL-R1), DCR1 (TRAIL-R3), DCR2 (TRAIL-R4), cFLIP or caspase-8 expression in either cell line (Figure 10a,b).

These data demonstrate that PMA promotes a more apoptotic equilibrium between functional and decoy TRAIL receptors, c-FLIP and caspase-8 in pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells by upregulating DR5 expression.

2.6. Jurkat Elimination of PMA-Activated TrkAIII-SH-SY5Y Cells Is Regulated by Trk, Shp, Src, IP3K and Akt

In live cell imaging assays, the enhanced sensitivity to Jurkat elimination exhibited by PMA-activated TrkAIII-SH-SY5Y cells, at a 2 to 1 ratio with NB cells, was significantly reduced by (i) the Trk inhibitor entrectinib [70] (1 μM, during PMA activation and present during assay) by 74.5% at 6 h, 45.1% at 12 h and 28.6% at 24 h (p < 0.0001 for all time points); (ii) the Shp inhibitor NSC-87877 [71] (1 μM, absent during PMA activation of TrkAIII-SH-SY5Y cells but present throughout the assay) by 100% at 6 h, 78.3% at 12 h and 33.9% at 24 h (p < 0.0001 for all time points); (iii) the Src inhibitor PP1 [72] (1 μM, absent during PMA activation of TrkAIII-SH-SY5Y cells but present throughout the assay) by 97.8% at 6 h, 74.9% at 12 h and 20.5% at 24 h (p < 0.0001 for all time points); (iv) the IP3K inhibitor LY294002 [73] (25 μM, absent during PMA activation of TrkAIII-SH-SY5Y cells but present throughout the assay) by 100% at 6 h, 80.9% at 12 h and 49.8% at 24 h (p < 0.0001 for all time points); and (v) by the Akt inhibitor capivasertib [74] (50 μM, absent during PMA activation of TrkAIII-SH-SY5Y cells but present throughout the assay) by 100% at 6 h, 62.4% at 12 h and 53.3% at 24 h (p < 0.0001 for all time points), but was not significantly inhibited by the ERK/MAPK inhibitor PD98059 [75] (10 μM, absent during PMA activation of TrkAIII-SH-SY5Y cells but present throughout the assay) (Figure 11).

These data implicate TrkAIII, Shp, Src, IP3K and Akt in the enhanced PMA-activated TrkAIII-SH-SY5Y sensitivity to Jurkat elimination.

2.7. Potential Evasion Mechanisms

2.7.1. SH-SY5Y Express Functional PD-L1

Constitutive pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y PD-L1 expression was confirmed by semi-quantitative RT-PCR (Figure 12a, upper 2 panels), Western blotting (Figure 12a, bottom 2 panels), qRT-PCR (Figure 12b), and also by indirect IF (Figure 12c). Significantly higher levels of TrkAIII-SH-SY5Y PD-L1 mRNA expression, detected by quantitative qRT-PCR, were significantly reduced by the Trk inhibitor entrectinib (1 μM for 24 h), which did not reduce pcDNA-SH-SY5Y PD-L1 mRNA expression. Enhanced PD-L1 immunoreactivity exhibited by TrkAIII-SH-SY5Y compared to pcDNA-S-SY5Y cells was also reduced following 48 h treatment with entrectinib (1 μM) (Figure 12c).

In IL-2 production assays [76], Jurkat IL-2 expression was induced by PMA/PHA but not by PMA alone, was abrogated in PMA/PHA-activated Jurkat cells by interaction with TrkAIII-SH-SY5Y cells, at a Jurkat to NB cell ratio of 1 to 8, and significantly restored by PD-L1 neutralizing antibody (5 μg/mL) [77], compared to pre-immune IgG (5 μg/mL) (Figure 13a and Supplementary Figure S2). Furthermore, IL-2 was not detected in 48 h-conditioned media from Jurkat co-cultures with either non-activated or PMA-activated TrkAIII-SH-SY5Y cells, at ratios of 1 to 1 (Supplementary Figure S2).

In contrast, in live cell imaging assays, the PD-L1 neutralizing antibody (5 μg/mL) did not significantly influence Jurkat cell elimination of PMA-activated TrkAIII-SH-SY5Y cells at either 6, 12 or 24 h, at a 2 to 1 ratio (Figure 13b).

These data confirm that SH-SY5Y cells express PD-L1, PD-L1 expression is augmented by the TrkAIII oncoprotein, SH-SY5Y PD-L1 is functional in blocking PMA/PHA-activated Jurkat IL-2 production, Jurkat interaction with PMA-activated TrkAIII-SH-SY5Y cells does not result in IL-2 production and the PD-L1/ PD-1 axis does not regulate Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells.

2.7.2. Osteoprotegerin (OPG) Is Expressed by SH-SY5Y Cells and Inhibits s-TRAIL-Induced but Not Jurkat m-TRAIL-Induced Elimination of PMA-Activated TrkAIII-SH-SY5Y Cells

Western blots detected the constitutive expression of the TRAIL inhibitor osteoprotegerin in whole cell extracts from pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells, which was not significantly altered by PMA treatment (60 ng/mL for 16 h) in either cell line (Figure 14a).

In live cell imaging assays, s-TRAIL (100 ng/mL) did not induce apoptosis in either non-activated or PMA-activated pcDNA-SH-SY5Y cells (Figure 14b) but induced apoptosis in non-activated TrkAIII-SH-SY5Y cells and significantly accelerated apoptosis in PMA-activated (60 ng/mL for 16 h) TrkAIII-SH-SY5Y cells at 6 h. Soluble recombinant OPG (1 μg/mL) completely abrogated s-TRAIL (100 ng/mL)-induced apoptosis of both non-activated and PMA-activated TrkAIII-SH-SY5Y cells (Figure 14c,d) but did not significantly reduce Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells at either 6, 12 or 24 h time points at a 2 to 1 ratio (Figure 14d,f).

These data indicate that SH-SY5Y cells express OPG, PMA does not sensitize pcDNA-SH-SY5Y cells to s-TRAIL, PMA enhances the sensitivity of TrkAIII-SH-SY5Y cells to s-TRAIL, and OPG prevents s-TRAIL but not Jurkat cell-mediated m-TRAIL-induced TrkAIII-SH-SY5Y apoptosis.

2.8. The Sensitivity of PMA-Activated NB Cells to Jurkat Elimination Is Associated with CCL-2 Expression and Is Also Regulated by Reciprocal NB Cell-Induced Jurkat Apoptosis

Analysis of NB cell sensitivity to Jurkat elimination was extended to PMA-activated parental SH-SY5Y, non-MYCN amplified SK-N-SH and MYCN-amplified SMS-KCNR, IMR-32 and NB-1 NB cell lines. Jurkat cells, at a 2 to 1 ratio to NB cells, eliminated PMA-activated parental SH-SY5Y and SK-N-SH cells with similar kinetics to PMA-activated pcDNA-SH-SY5Y cells (Figure 15a).

Jurkat cells did not eliminate non-activated SMS-KCNR, IMR-32 or NB-1 cells but induced delayed elimination of PMA-activated (60 ng/mL for 16 h followed by removal) SMS-KCNR cells at a 6 to 1 ratio to NB cells, which was not detected prior to 48 h. In contrast, Jurkat cells did not eliminate either PMA-activated IMR-32 or NB-1 cells at a 6 to 1 ratio with NB cells over the same time course (Figure 15b).

RT-PCRs confirmed PMA (60 ng/mL for 6 h) induction of CCL2 mRNA expression in SMS-KCNR but not in either NB-1 or IMR-32 cells (Figure 15d), and Western blots confirmed the induction of CCL2 compared to MMP-2 protein expression in 20-fold concentrated 48 h serum-free conditioned media from PMA-activated (60 ng/mL for 16 h followed by removal and 48 h serum-free culture) SMS-KCNR cells compared to non-activated SMS-KCNR counterparts (Figure 15e).

Western blots also detected caspase-3 cleavage in whole cell extracts from Jurkat cells isolated after a 24 h interaction with non-activated pcDNA-SH-SY5Y, TrkAIII-SH-SY5Y, IMR-32 and SMS-KCNR cells at a 4 to 1 ratio with NB cells. In contrast, caspase-3 cleavage was not detected in Jurkat cells isolated following a 24 h interaction with non-activated IMR-32 or SMS-KCNR cells in the presence of MET-12 (10 μM) (Figure 15f).

These data demonstrate that at high Jurkat to NB cell ratios, Jurkat cells can also eliminate PMA-activated MYCN-amplified SMS-KCNR cells that exhibit PMA-induced CCL2 expression but not PMA-activated MYCN-amplified IMR-32 and NB-1 cells, which exhibit repression of PMA-induced CCL2 expression. The data also indicate that NB cells exhibit reciprocal Fas/Fas-induced apoptosis of Jurkat cells.

3. Discussion

With the aim of identifying alternative and/or complementary ways to harness the immune system for the treatment of high-risk non-antigenic, non-immunogenic NBs, we present the characterization of an antigen-independent mechanism for eliminating non-MYCN-amplified (pcDNA-SH-SY5Y, TrkAIII-SH-SY5Y and parental SH-SY5Y) and MYCN-amplified (SMS-KCNR) NB cells in a Jurkat model of lymphocyte cytotoxicity. This mechanism does not require Jurkat T-cell pre-activation but depends upon PMA activation of NB cells. It is initiated by the induction of NB cell CCL2 chemokine expression, CCL2 promotion of Jurkat CCR2b expression and subsequent CCL2/CCR2-mediated, inside-out activation of Jurkat LFA-1. This results in the transition to Jurkat LFA-1-dependent adhesion and the formation of antigen-independent, cytotoxic immune synapses between Jurkat LFA-1 and NB cell ICAM-1, through which mobilized Jurkat TRAIL vesicles combine with enhanced NB cell DR5 expression to induce cell-mediated m-TRAIL-mediated NB cell apoptosis via the extrinsic pathway. This cytotoxic mechanism is enhanced by the expression of the NB-associated TrkAIII oncoprotein via cFLIP sequestration, is PD-L1/PD-1 independent, is resistant to OPG and eliminates PMA-activated non-MYCN-amplified (parental SH-SY5Y, pcDNA-SH-SY5Y, TrkAIII-SH-SY5Y and SK-N-SH) and MYCN-amplified (SMS-KCNR) NB cells that express CCL2 but not MYCN-amplified NB cells (IMR-32 and NB-1) that exhibit CCL2 repression, and is offset by variable NB cell FasL/Fas-induced Jurkat cell apoptosis.

PMA sensitization of SH-SY5Y cells to elimination by Jurkat cells was first detected while investigating TrkAIII-SH-SY5Y inhibition of IL-2 production by PMA/PHA-activated Jurkat cells. This revealed that both low and high TrkAIII-expressing SH-SY5Y NB cells, resistant to Jurkat elimination under non-activated conditions, were sensitized by PMA to elimination by non-pre-activated Jurkat cells, implicating PMA-induced PKC activation and downstream signaling pathways [78,79]. Focusing on the enhanced sensitivity of TrkAIII-SH-SY5Y cells, Jurkat cytotoxicity was significantly reduced by LY-294002 and capivasertib but not by PD98059, implicating PKC-induced IP3K and Akt but not MAPK signaling [78,79]. PMA also suppressed pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y proliferation, previously attributed to PKC-induced AMPK signaling in SH-SY5Y cells [80]. With respect to PMA use in cancer research, this tumor-promoting agent is a high-affinity broad activator of PKC and downstream signaling, and represents a potent pharmacological equivalent to natural PKC activators in the tumor microenvironment [81,82,83,84]. In NB research, PMA has been used to stimulate or amplify signaling pathways that regulate the expression of genes that promote tumor progression but also reduces NB cell proliferation and promotes neuronal-like differentiation [80,85,86], consistent with a potential tumor-suppressing role for PKC activation in different cancers [87]. Therefore, the novel role for PKC activation induced by PMA in sensitizing NB cells to T-cell elimination, reported in this study, unveils a potentially exploitable immunological “Achilles heel”, supporting the concept that PKC activation may be a valuable therapeutic strategy in NB. Jurkat elimination of PMA-activated pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells was associated with Jurkat recruitment and adherence to NB cell surfaces and was not transferred in cell-free conditioned media from Jurkat cell co-cultures with PMA-activated TrkAIII-SH-SY5Y cells, collected at times of maximal NB cell death. This confirms a cell-mediated cytotoxic mechanism. Cytotoxicity was abrogated by the TRAIL neutralizing antibody [59] but not by the small peptide Fas inhibitor MET-12 [60] or by the Htr2A/Omi inhibitor UCF-101 [64]. This implicates a mechanism largely dependent upon Jurkat membrane-bound m-TRAIL rather than Granzyme B or FasL, the latter also supported by a report that SH-SY5Y cells do not express Fas [88].

LFA-1 and ICAM-1 are involved in T-cell cytotoxicity [89,90,91,92,93], Jurkat cells express the LFA-1 subunit CD11a, and SH-SY5Y cells express the LFA-1 ligand ICAM-1. The LFA-1 inhibitor BIRT377 [65] and recombinant ICAM-1 inhibited Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells, confirming roles for both in this mechanism. However, PMA did not enhance ICAM-1 expression in either pcDNA-SH-SY5Y or TrkAIII-SH-SY5Y cells, implicating additional factors in sensitization to elimination. In this regard, BIRT377 inhibited Jurkat cell adhesion to PMA-activated but not to non-activated TrkAIII-SH-SY5Y cells, associating cytotoxicity with a transition from Jurkat LFA-1-independent to LFA-1-dependent adhesion. Jurkat cells express inactive LFA-1 (this study) [90], implicating a PMA-induced, NB cell-initiated, LFA-1 integrin-activating chemokine/chemokine receptor axis [91,92,93]. In support of this, PMA promoted expression and secretion of the integrin-activating chemokine CCL2 in pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells. Furthermore, Jurkat cell interaction with PMA-activated but not non-activated TrkAIII SH-SY5Y cells enhanced Jurkat CCL2 chemokine receptor CCR2b expression. The CCL2 inhibitor bindarit [66] reduced CCL2 expression by PMA-activated TrkAIII-SH-SY5Y cells to levels detected in non-activated counterparts and also prevented the induction of Jurkat CCR2b mRNA expression following interaction with PMA-activated TrkAIII-SH-SY5Y cells. Furthermore, both bindarit and the CCR2 inhibitor INCB3284 [67] prevented Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells. This strongly suggests that this cytotoxic mechanism is CCL2/CCR2-dependent, which is also supported by reports that CCL2 expression is promoted by NF-κB activators in SH-SY5Y cells and other cell types, which would include PMA [94,95] and a report that CCL2 promotes CCR2 expression [96]. We consider that limited constitutive CCL2 expression by non-activated TrkAIII-SH-SY5Y cells was below the threshold required to promote Jurkat CCR2b expression, LFA-1 activation and cytotoxicity.

With respect to alternative chemokine axes, Jurkat cells also express the CXCL12 receptor CXCR4 (this study) [97,98], and pcDNA-SH-SY5Y cells express the integrin-activating chemokine CXCL12. However, CXCL12 mRNA expression was not detected in either non-activated or PMA-activated TrkAIII-SH-SY5Y cells. Furthermore, the interaction of Jurkat cells with PMA-activated TrkAIII-SH-SY5Y cells led to a marked reduction in CXCR4 expression, and the CXCR4 inhibitor IT1t-dihydrochloride [68] did not significantly reduce Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells, which would exclude CXCL12/CXCR4 involvement in this mechanism, potentially also influenced by CCL2/CCR2 de-sensitization of cells to the CXCL12/CXCR4 axis [98]. Bindarit did not restore Jurkat cell CXCR4 expression following interaction with PMA-activated TrkAIII-SH-SY5Y cells, excluding CCL2 involvement. The molecular mechanism involved in TrkAIII repression of CXCL12 expression is under further investigation.

LFA-1 and ICAM-1 involvement in this cytotoxic mechanism was associated with IF evidence of supramolecular activation cluster (SMAC) formation at sites of Jurkat interaction with PMA-activated but not non-activated TrkAIII-SH-SY5Y cells. SMACs were characterized by the clustering of Jurkat CD11a, and SH-SY5Y ICAM-1 and DR5/TRAIL-R2 with ring formation, re-localization of TrkAIII-SH-SY5Y Golgi (GM130) and centrosomes (γ-tubulin) to ICAM-1 and DR5 clusters, and polarization of Jurkat TRAIL vesicles towards interaction sites, all of which are consistent with antigen-independent LFA-1/ICAM-1 immune synapse formation [92,99,100]. In addition, the lipid-raft disrupting agent methyl-β-cyclodextrin [69] also prevented Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells, confirming a requirement for lipid rafts, which are fundamental for the formation of immune synapses and canonical TRAIL/TRAIL receptor apoptotic signaling [100,101,102]. Cytotoxic LFA-1/ICAM-1 immune synapse formation also helps to explain the inhibitor-confirmed involvement of IP3K and Akt, since the IP3K/Akt axis regulates immune synapse formation and function [103]. Jurkat m-TRAIL-induced NB cell apoptosis also depends on a pro-apoptotic equilibrium between functional and decoy TRAIL receptors, c-FLIP and caspase-8 [35,104]. In this regard, PMA increased functional TRAIL receptor DR5 (TRAIL-R2) expression in pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells without significantly altering the expression of DR4 (TRAIL-R1), DCR1 (TRAIL-R3), DCR2 (TRAIL-R4), c-FLIP or caspase-8, consistent with a shift to an apoptotic equilibrium.

Jurkat cells also express TCRs of unknown specificity [51] and SH-SY5Y cells exhibit MHC class I/II protein expression [105,106]. Therefore, the potential involvement of noncanonical antigen-independent interactions between SH-SY5Y MHCs and Jurkat TCRs cannot be excluded. However, PD-L1/PD-1 inhibits low-affinity TCR signaling [107,108], whereas the neutralizing PD-L1 antibody did not prevent Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells. Furthermore, Jurkat cells did not produce IL-2 upon interaction with PMA-activated TrkAIII-SH-SY5Y cells, which is regarded as a sensitive and quantitative index of Jurkat TCR signaling [109]. These observations suggest that any potential involvement of antigen-independent non-canonical MHC class I/II-TCR interactions in this mechanism would be limited to synapse formation in the absence of TCR activation and signaling. Jurkat cells also efficiently eliminated PMA-activated TrkAIII-SH-SY5Y cells at low ratios, consistent with short-term cytotoxic immune synapse engagement, which optimizes synapse maturation and activation for targeted TRAIL release and killing without compromising detachment and engagement of other targets. This may also reflect antigen-independent cytotoxicity, since continuous exposure to antigens promotes prolonged immune synapse engagement, preventing detachment and engagement of other targets, causing cytotoxic paralysis and T-cell exhaustion [110,111,112].

PMA-activated TrkAIII-SH-SY5Y cells were significantly more sensitive to Jurkat elimination than PMA-activated pcDNA-SH-SY5Y cells, extending our previous report that the TrkAIII oncoprotein sensitizes SH-SY5Y cells to s-TRAIL-induced apoptosis [29], to enhanced sensitivity to Jurkat m-TRAIL-induced apoptosis. This enhanced sensitivity was reduced by Trk (entrectinib), Shp (NSC87877) and Src (PP1) inhibitors, previously implicated in TrkAIII sensitization of SH-SY5Y cells to s-TRAIL-induced apoptosis, implicating the same mechanism, which has been previously shown to depend upon TRAIL-induced, Shp/Src-regulated TrkAIII sequestration of c-FLIP [29]. This confirms that cFLIP availability is a critical determinant of NB cell sensitivity to s-TRAIL and m-TRAIL-induced apoptosis and also explains the relative resistance of PMA-activated pcDNA-SH-SY5Y cells to s-TRAIL, despite similar levels of DR5, CCL2, CD11a and ICAM-1 expression. The fact that pcDNA-SH-SY5Y resistance to Jurkat elimination could be overcome by increasing the Jurkat to NB cell ratio indicates that the inhibitory cFLIP threshold can be surmounted by increasing Jurkat numbers and, by inference, immune synapse numbers. Conversely, the complete resistance exhibited by PMA-activated pcDNA-SH-SY5Y cells to s-TRAIL indicates that c-FLIP more readily prevents s-TRAIL compared to m-TRAIL-induced apoptosis, which most likely reflects the enhanced potency of membrane-stabilized m-TRAIL trimers, oligomerized with TRAIL receptors within the context of lipid-raft immune synapses [100,101,113]. On the other hand, the enhanced PMA-activated TrkAIII-SH-SY5Y sensitivity to s-TRAIL can be explained by the combination of reduced c-FLIP availability [29] and enhanced DR5 expression, whereas non-activated TrkAIII-SH-SY5Y resistance to Jurkat elimination can be explained by LFA-1-independent adhesion, absence of LFA1/ICAM-1 immune synapses and TRAIL vesicle polarization associated with the absence of cell surface m-TRAIL expression on non-activated Jurkat cells.

Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells was associated with caspase-8 and caspase-3 cleavage and was prevented by z-VAD-fmk (pan caspase inhibitor), z-IEDT-fmk (caspase-8 inhibitor) and z-DEVD-fmk (caspase-3 inhibitor), confirming activation of the extrinsic apoptosis pathway. Although TRAIL can also activate the intrinsic mitochondrial pathway [114], the Omi/HTR2A inhibitor UCF-101 did not reduce cytotoxicity, which not only excludes the intrinsic mitochondrial pathway [115,116] but also the involvement of granzyme B, which activates mitochondrial apoptosis via BID cleavage [117], supporting a mechanism dependent upon the differential mobilization of Jurkat TRAIL vesicles [54,118].

With regard to ways, other than enhanced c-FLIP availability, to evade this mechanism, the potential involvement of the PD-1/PD-L1 immune checkpoint [119] was evaluated. PD-L1 expression was detected in pcDNA-SH-SY5Y cells and was significantly augmented in TrkAIII-SH-SY5Y cells, adding the promotion of PD-L1 expression to TrkAIII’s oncogenic repertoire. This, combined with PD-1 expression in Jurkat cells [76], identifies PD-L1/PD-1 as a potential evasion mechanism. The neutralizing PD-L1 antibody confirmed TrkAIII-SH-SY5Y PDL1 function in PMA/PHA-activated Jurkat IL-2 production assays, but failed to influence Jurkat elimination of PMA-activated TrkAIII-SH-SY5Y cells, confirming that this cytotoxic mechanism is PD-L1/PD-1-independent. This also indicates that cytotoxicity does not depend upon full Jurkat PKC/Ca^2+^ activation (this study) or the mobilization of Granzyme B granules, both of which are inhibited by PD-L1/PD-1 [76,120].

TrkAIII-SH-SY5Y and pcDNA-SH-SY5Y cells also express the TRAIL inhibitor OPG [42], and recombinant OPG abrogated s-TRAIL-induced apoptosis of both non-activated and PMA-activated TrkAIII-SH-SY5Y cells, unveiling an additional potential evasion mechanism. However, OPG did not significantly reduce Jurkat m-TRAIL-induced elimination of PMA-activated TrkAIII-SH-SY5Y cells. This indicates that OPG inhibition of TRAIL does not necessarily extend to m-TRAIL, especially within this LFA-1/ICAM-1 immune synapse context. Whether this reflects difficulty in permeating the tight junctions that form in mature LFA-1/ICAM-1 immune synapses [121], or a more efficient capacity of OPG to destabilize s-TRAIL trimers [122], compared to highly stabilized m-TRAIL trimers oligomerized with TRAIL receptors within immune synapses [102,105,113], remains to be elucidated. However, this has important potential implications for the use of s-TRAIL or cell-mediated m-TRAIL immunotherapeutic strategies for the treatment of NBs that express OPG and suggests that cell-mediated m-TRAIL delivery could be combined with OPG for the treatment of osteolytic NB bone metastases [45,122].

Jurkat cells, in addition to eliminating PMA-activated pcDNA-SH-SY5Y and TrkAIII-SH-SY5Y cells, also eliminated PMA-activated parental SH-SY5Y cells and non-MYCN-amplified SK-N-SH cells with similar kinetics to pcDNA-SH-SY5Y cells, at a 2 to 1 ratio with NB cells. In addition, Jurkat cells also induced delayed (48 to 72 h) elimination of PMA-activated MYCN-amplified SMS-KCNR NB cells at a 6 to 1 ratio, extending the cytotoxicity mechanism to this MYCN-expressing MYCN-amplified NB cell line [123]. In contrast, Jurkat cells did not eliminate either PMA-activated MYCN-expressing MYCN-amplified IMR-32 or NB-1 NB cells [124,125], even at a 6 to 1 Jurkat to NB cell ratio. In potential explanation for this, PMA induced CCL2 expression in Jurkat-sensitive pcDNA-SH-SY5Y, TrkAIII-SH-SY5Y and SMS-KCNR cells but not in Jurkat-resistant IMR-32 and NB-1 cells, corroborating the hypothesis that CCL2 is a critical component of this cytotoxic mechanism and identifying repression of CCL2 expression in MYCN-amplified NB-1 and IMR-32 cells as a potential evasion mechanism. These observations add to reports that MYCN represses NB cell CCL2 production, lymphocyte recruitment and cytotoxicity [126,127,128] but reveal that CCL2 repression does not characterize all MYCN-amplified NB cell lines. In addition, caspase-3 cleavage was detected in Jurkat cells isolated following interaction with non-activated pcDNA-SH-SY5Y, TrkAIII-SH-SY5Y, SMS-KCNR, and IMR-32 cells at Jurkat to NB cell ratios equal to or greater than 4 to 1, which was prevented by MET12 in Jurkat cells isolated after a 24 h interaction with SMS-KCNR and IMR-32 cells, implicating Fas-induced Jurkat cell apoptosis. This supports reports that NB cell lines express FasL and that FasL induces Jurkat apoptosis [88,129], and indicates that the Jurkat to NB cell ratio required for predominant elimination of CCL2-expressing NB cells also depends upon the reciprocal rate of NB cell FasL/Fas-induced Jurkat cell apoptosis.

Although NBs are generally considered to be immunologically “cold”, they are highly heterogeneous, exhibit context-dependent differences in T-cell and myeloid cell infiltration, and are characterized by more differentiated, less immunogenic adrenergic components and less differentiated, more immunogenic mesenchymal stem cell-like components, which exhibit differences in MHC and cytokine expression. Adrenergic and mesenchymal phenotype plasticity is regulated by the tumor microenvironment, and mesenchymal conversion is promoted by chemotherapeutic agents and is involved in drug resistance, post-therapeutic relapse and metastatic progression [14,130]. Drug-induced mesenchymal conversion, therefore, unveils a potential immunological “Achilles heel” that could be exploited by combining conventional chemotherapy with immunotherapy [14]. Although phenotypic heterogeneity was not investigated in this study, TrkAIII-SH-SY5Y cells exhibit a more mesenchymal stem cell-like phenotype [131], which may also help to explain enhanced sensitivity to this cytotoxic mechanism.

This pre-clinical basic research study, using the Jurkat leukemia T-cell line as a T-cell surrogate, currently has limited clinical value. It does, however, reveal a unique T-cell-mediated antigen and TCR activation-independent mechanism for eliminating PKC-activated NB cells. This calls for the identification of clinically relevant physiological immune cells that exhibit this mechanism. Given the fundamental requirements of TRAIL, LFA-1, and CCR2b expression for cytotoxicity, possible physiological immune cell equivalents include CD8+, CD4+ and γ/δ T cells, as well as NK cells, which, despite similarities and differences in death receptor and MHC-dependent cytotoxicity, have all been demonstrated to express TRAIL, LFA-1, and CCR2b [132,133,134,135,136,137,138,139,140,141,142,143].

Although PKCs are generally considered oncogenic kinases, they also act as tumor suppressors in different cancers [87] including NB [80]. In this study, PMA reduced NB cell proliferation and also sensitized NB cells to elimination by Jurkat T cells, suggesting that PKC activation may be a valuable therapeutic strategy in NB. PMA, however, is not a clinically approved PKC activator; therefore, more research will be required to assess whether the clinically approved PKC activators bryostatin-1, ingenol mebutate or tigilanol tiglate [144,145,146], or specific activators of PKC isoforms, diacylglycerol-lactones, thymeleatoxin, HMI-1b11 or 7α-acetoxy-6β-benzoyloxy-12-O-benzoylroyleanone [147,148,149,150], currently under development, may also sensitize NB cells to this cytotoxic mechanism. Research will also be required to ascertain how metabolic acidosis, nutrient deprivation, hypoxia, inflammation and chemotherapy within the tumor microenvironment [151] may influence this mechanism.

4. Materials and Methods

4.1. Cell Lines and Culture Conditions

Parental SH-SY5Y, SK-N-SH and SMS-KCNR NB cell lines were obtained from DR. U.P. Thorgeirsson (NCI, NIH Bethesda, MD, USA). Parental SH-SY5Y cells were used to obtain stable transfected pcDNA and TrkAIII cell lines [21]. Jurkat (E6.1) and IMR-32 NB cell lines were from ATCC (Manassas, VI, USA) and the NB-1 cell line (RCB1953) was from RIKEN BRC (Tsukuba, Japan). NB and Jurkat cell lines were cultured in RPMI 1640, supplemented with 10% fetal bovine serum, 1% glutamine, and 1% penicillin/streptomycin (Euroclone, Milan, Italy). Stable pcDNA, TrkA, and TrkAIII-SH-SY5Y transfectants were cultured intermittently in complete medium containing zeocin (200 μg/mL) (Thermo Fisher Scientific, Waltham, MA, USA).

4.2. Reagents and Antibodies

LY-290024, PD98059, phytohemagglutinin (PHA), 12-O-tetradecanoyl-phorbol-D-acetate (PMA) and methyl-β-cyclodextrin were from Sigma-Aldrich (Saint Louis, MO, USA). Entrectinib was from Selleck Chemicals (Houston, TX, USA). Zeocin was from Invitrogen (Carlsbad, CA, USA). UCF-101, z-VAD-fmk, z-IETD-fmk, z-DEVD-fmk, INCB3284, Capivasertib, NSC-87877, BIRT377, PP1, MET12, soluble recombinant TRAIL (s-TRAIL), recombinant human ICAM-1 (rICAM-1), IT1t-dihydrochloride, Bindarit and recombinant human osteoprotegerin were from MedChemExpress (Monmouth Junction, NJ, USA). Mouse monoclonal anti-human PD-L1 (1C10, sc-293425), mouse monoclonal anti-human osteoprotegerin (E-10, sc-390518), mouse monoclonal anti-human beta-actin (C4, sc-47778), rabbit polyclonal anti-human GM130 (H65, sc-30100), rabbit polyclonal anti-human c-FLIP (H-150, sc-8346) and mouse monoclonal anti-human gamma-tubulin (D-10, sc-17788) antibodies were from SantaCruz (Santa Cruz, CA, USA). Rabbit monoclonal anti-human caspase-3 (14220), rabbit monoclonal anti-human cleaved caspase-3 (9664), mouse monoclonal anti-human caspase-8 (9746), rabbit polyclonal anti-human caspase-9 (9502), rabbit monoclonal anti-human CCL2 (F9I9N) and rabbit monoclonal anti-human CCR2 (D14H7) antibodies were from Cell Signaling Technology (Danvers, MA, USA). Rabbit polyclonal anti-human ICAM-1(CSB-PA07149A0Rb) and rabbit polyclonal anti-human CD11a (ITGAL CSB-PA011875LA01HU) antibodies were from Cusabio (Houston, TX, USA). Rabbit polyclonal anti-human TRAIL (54008), DR4 (54024), DR5 (28065), DCR1 (54053), and DCR2 (54036) antibodies were from AnaSpec (Fremont, CA, USA). Rabbit neutralizing recombinant monoclonal anti-human PD-L1 (14-5983-82) (77) and mouse neutralizing recombinant monoclonal anti-human TRAIL (16-9927-82) antibodies were from Invitrogen (ThermoFisher Scientific, Waltham, MA, USA). Rabbit anti-human MMP-2 polyclonal antibody was produced in our laboratory [152]. Secondary horseradish peroxidase (HRP)-conjugated antibodies were from Bethyl Laboratories Inc. (Fortis, Waltham, MA, USA). Alexa Fluor 488-labeled donkey anti-mouse and Alexa Fluor 594-labeled donkey anti-rabbit antibodies were from Life Technologies (Waltham, MA, USA). The ProLong^TM^ Gold anti-fade reagent with DAPI was from Invitrogen (ThermoFisher Scientific, Waltham, MA, USA).

4.3. Protein Extraction and Western Blotting

For whole cell protein extraction, cells were lysed in RIPA buffer (1× PBS, 0.5% sodium deoxycholate, 1% NP40, 0.1% SDS, 1 mM sodium orthovanadate, 1 mM PMSF, 1 μg/mL of pepstatin A and Aprotinin) and protein concentrations determined by Bradford assay (Sigma-Aldrich). For Western blotting, protein extracts in 1 × SDS-PAGE reducing buffer (4% SDS, 5% 2-mercaptoethanol, 20% glycerol, 0.004% bromophenol blue and 0.125 M Tris-HCl (pH 6.8)) were heated to 95 °C for 5 min, quenched on ice, then subjected to reducing SDS-PAGE. Following electrophoresis, proteins were transferred onto Hybond C^+^ nitrocellulose membranes (Amersham Int., Bedford, UK) by electrophoresis, and membranes were air-dried. Non-specific binding sites on air-dried membranes were blocked by a 1 h incubation in TBS-T blocking solution (15 mM NaCl, 10 mM Tris-HCl pH 8.0, 10% Tween 20), containing 5% low-fat milk powder (Euroclone), then incubated overnight with appropriate dilutions of primary antibodies in blocking solution at 4 °C with rotation. Membranes were then washed in TBS-T solution and incubated with appropriate dilutions of horseradish peroxidase (HRP)-conjugated secondary antibodies (Bethyl Laboratories Inc.) for 1 h in blocking solution at room temperature. Immunoreactive species were detected using an ECL chemiluminescence kit, as directed by the manufacturer (Amersham Int, Bedford, UK), in an ImageQuant™ LAS 4000 (GE Healthcare) image analyzer. Where stated, Western blots were subjected to ImageJ pixel densitometric analysis (http://imagej.nih.gov/ij/ FIJI version, accessed on 10 November 2025).

4.4. RNA Extraction, Semiquantitative RT-PCR and Quantitative qRT-PCR

Total cellular RNAs (1 µg), purified from cell cultures using a Quick-RNA™ Miniprep Kit, as described by the manufacturer (Zymo Research, Freiberg im Breisgau, Germany), were reverse-transcribed using SuperScript™ IV VILO™ Master Mix, as described by the manufacturer (Thermo Fischer Scientific, Waltham, MA, USA). For semi-quantitative RT-PCRs, linear sub-plateau phase amplification was established by comparing cDNA serial dilutions ranging from non-diluted to 1 to 100,000, for each individual gene and primer set, in 35-cycle reactions [153]. The final cDNA dilutions and volumes used were as follows: 1 μL of 1 to 100,000 dilution (0.0005 ng cDNA) for 18S rRNA; 1 μL non-diluted (50 ng cDNA) for CCR2 and CCL2; and 1 μL of a 1 to 10 dilution (5 ng cDNA) for PD-L1. SYBR green quantitative qRT-PCR was performed using Luna^®^ Universal qPCR Master Mix, as described by the manufacturer (New England Biolabs, Ipswich, MA, USA) and evaluated in an Applied Biosystems^®^ 7500 Real-Time PCR System (Thermo Fischer Scientific, Waltham, MA, USA). The cDNA volumes and dilutions used in qRT-PCRs were as follows: 1 µL of a 1 to 10 dilution (5 ng cDNA) for PDL1. Relative quantification was performed using the real-time PCR 2(-Delta Delta C(T)) method, normalized for 18S rRNA, as previously described [154]. The following primers (and PCR conditions) were used: 18S rRNA: 5′-AAACGGCTACCACATCCAAG-3′ and 5′-CCTCGAAAGAGTCCTGTATTG-3′ (denaturation 30 s at 94 °C, annealing 30 s at 58 °C and extension 30 s at 72 °C); PD-L1: 5′-GTGGCATCCAAGATACAAACTCAA-3′ and 5′-TCCTTCCTCTTGTCACGCTCA-3′ (denaturation 30 s at 94 °C, annealing 30 s at 58 °C and extension 30 s at 72 °C); CCL2: 5′-CAGGTGACAGAGACTCTTGGGA-3′ and 5′-GGCAATCCTACAGCCAAGAGCT-3′ (denaturation 30 s at 94 °C, annealing 30 s at 60 °C and extension 30 s at 72 °C); CCR2: 5′-CAGGTGACAGAGACTCTTGGGA-3′ and 5′-GGCAATCCTACAGCCAAGAGCT-3′ (denaturation 30 s at 94 °C, annealing 30 s at 60 °C and extension 30 s at 72 °C); CXCR4: 5′-AGCATGACGGACAAGTACAGG-3′ and 5′-GATGAAGTCGGGAATAGTCAG-3′ (denaturation 30 s at 94 °C, annealing 30 s at 53 °C and extension 30 s at 72 °C); CXCL12: 5′-CTCAACACTCCAAACTGTGCCC-3′ and 5′-CTCCAGGTACTCCTGAATCCAC-3′ (denaturation 30 s at 94 °C, annealing 30 s at 58 °C and extension 30 s at 72 °C). All RT-PCRs and RT-qPCRs were performed in triplicate and repeated a minimum of twice. Where stated, gels containing RT-PCR products were subjected to ImageJ pixel densitometric analysis (http://imagej.nih.gov/ij/ FIJI version, accessed on 20 November 2025).

4.5. Indirect Immunofluorescence

TrkAIII-SH-SY5Y cells grown to sub-confluence on Nunc glass chamber slides (Sigma-Aldrich, St. Louis, MO, USA) were treated for 16 h in the presence or absence of PMA (60 ng/mL in complete medium), extensively washed in pre-warmed serum-free RPMI (37 °C) to remove traces of PMA, then co-cultured for 10 h in complete medium with non-pre-activated Jurkat cells at a Jurkat to NB cell ratio of 2 to 1. Co-cultures were then rinsed in pre-warmed PBS (37 °C), fixed in 10% formalin (v/v) and permeabilized in 100% ice-cold methanol at −20 °C. Fixed, permeabilized cells were incubated for 1 h in blocking solution (1% bovine serum albumin in PBS containing 0.03% Triton X-100), then incubated for 2 h with appropriate primary antibody dilutions in blocking solution at room temperature, then washed in PBS and incubated for 1 h at room temperature with appropriate dilutions of Alexa Fluor-conjugated secondary antibodies in blocking solution. Following incubation, slides were washed with PBS and mounted with ProLong™ Gold Antifade reagent, containing DAPI for nuclear staining (Thermofisher Scientific, Waltham, MA, USA). For analysis of cell surface and intracellular TRAIL expression, non-activated Jurkat cells were seeded onto Nunc glass chamber slides in the absence of serum and allowed to adhere, then either fixed and permeabilized or processed as live non-permeabilized cells by indirect IF for anti-TRAIL immunoreactivity (described above). Immunoreactivity was observed and documented by micrography, under a Zeiss Axioplan 2 fluorescent microscope (Oberkochen, Germany), equipped with a digital camera and Leica M500 Image Manager software.

4.6. Adhesion Assays

Using a modification of a previously described adhesion assay [155], TrkAIII-SH-SY5Y cells were grown to sub-confluence on Nunc glass chamber slides (Sigma-Aldrich, St. Louis, MO, USA) and incubated for 16 h with or without PMA (60 ng/mL, in complete medium). At 16 h, Jurkat cells were pre-incubated in the presence or absence of BIRT377 (15 μM for 30 min, followed by removal), then added to PMA-activated or non-activated TrkAIII-SH-SY5Y cells at a concentration of 1 × 10^6^ cells/mL. The cells were incubated for 60 min at 37 °C. At 60 min, the slides were de-chambered and passed 10 times through a pre-warmed (37 °C) air/PBS interface to remove non-adherent cells. The cells were fixed in a 96% ethanol–3% glacial acetic acid solution for 20 min at 4 °C, stained with hematoxylin and eosin, dehydrated, cleared in xylene and mounted with DePeX mounting medium (Sigma-Aldrich, St. Louis, MO, USA). Negative image micrographs of Jurkat cells adherent to TrkAIII-SH-SY5Y monolayers were documented under a Zeiss Axioplan 2 microscope, equipped with a digital camera and Leica M500 Image Manager software. The numbers of Jurkat cells adherent to TrkAIII-SH-SY5Y cells was quantified by direct counting of micrographs. Experiments were performed in duplicate and repeated 3 times.

4.7. Live-Cell Imaging Analysis of Cell Death

NB cells, seeded at a concentration of 1.5 × 10^4^ cells/well in 96-well culture plates (353072, Corning Inc., New York, NY, USA), were allowed to adhere and were pre-incubated for 16 h in the presence or absence of PMA (60 ng/mL). Following pre-incubation, cultures were washed 3 times with 200 uL of pre-warmed PBS (37 °C) to eliminate residual PMA, and Jurkat cells were added at different ratios (Results Section), in the presence or absence of inhibitors or neutralizing antibodies (Results Section). Live-cell time-lapse imaging was performed in an IncuCyte^®^ S3 Live-Cell Analysis System incubator, according to the manufacturer’s instructions (Sartorius, Goettingen, Germany), and images of two independent areas per well were recorded at 3 h intervals, at 10x magnification. Cell behavior was then analyzed in time-lapse videos. NB cell growth and Jurkat elimination of NB cells were quantified by direct counting of live and overtly dead SH-SY5Y cells in 3 fields per captured video image at the time points described in the Results section. Experiments were performed in duplicate and repeated a minimum of 3 times.

4.8. IL-2 ELISA

IL-2 production was quantified by IL-2 ELISA, as described by the manufacturer (Thermo Fischer Scientific, Waltham, MA, USA), in 48 h-conditioned media from non-activated and PHA/PMA-activated Jurkat cells (1 μg/mL PHA and 60 ng/mL PMA), non-activated TrkAIII-SH-SY5Y cells, and in co-cultures of PHA/PMA-activated Jurkat cells and TrkAIII-SH-SY5Y cells, in the presence or absence of pre-immune IgG (10 μg/mL) or neutralizing anti-PD-L1 antibody (5 µg/mL), at a Jurkat to NB cell ratio of 1 to 8, as previously described [76]. IL-2 ELISAs were quantified in an ELISA plate reader (Infinite M Plex, Tecan, Männedorf, Swiss), at 450 nm. Experiments were performed in triplicate and repeated a minimum of 2 times.

4.9. Statistical Analysis and Software

Data were analyzed by Student’s t-test or one-way ANOVA, using the online t-test calculator at https://www.graphpad.com/quickcalcs/ttest1.cfm (accessed 20 November 2025), with significant differences associated with probabilities of ≤0.05.

5. Conclusions

Although the CCL2/CCR2 axis is considered a clinical target due to its role in promoting monocyte recruitment, monocyte-induced angiogenesis and tumor-promoting inflammation [156,157,158,159], this axis also regulates tumor infiltration by antigen-presenting dendritic cells and cytotoxic lymphocyte populations. Furthermore, it has been implicated in dendritic cell-prone microenvironments and improved outcomes in low-risk non-MYCN amplified NBs and is repressed in high-risk MYCN-amplified NBs that exhibit poor dendritic and T-cell infiltration and outcomes. This identifies a potential therapeutic use for CCL2 in promoting dendritic and T cell infiltration and improving outcomes [160]. Moreover, CCL2-expressing tumors, which include ≈50% of metastatic high-risk NBs, show enhanced T-cell surveillance, NK infiltration, and responses to adoptive T-cell, CCR2b-expressing CAR T-cell and GD2 CAR T-cell immunotherapy [129,161,162,163,164,165]. Thus, the Jurkat T-cell antigen and TCR-activation-independent CCL2/CCR2-mediated cytotoxic mechanism described in this study provides a solid foundation for the future identification of clinically relevant physiological immune equivalents that exhibit this mechanism, validation of alternative PKC activators, and eventual assessment of the impact of tumor microenvironmental conditions. The ultimate goal of these studies is to assess the feasibility of translating this cytotoxic PKC activation-mediated mechanism into an effective immune-therapeutic approach for the treatment of high-risk NBs, especially NBs that exhibit CCL2 and TrkAIII expression.