The detailed analysis of the phenotypic and functional characterization of γδ T cells present in CRC can provide a scenario to understand the biological significance of this lymphocyte subset in this disease. This analysis can be performed by immunohistochemistry (IHC), flow cytometry, and molecular single-cell mRNA techniques [11, 48–54]. The cross-talk of γδ T cells with the diverse components of the tumor microenvironment (TME) is the context in which the engagement of specific receptors on γδ T cells may lead to pro-tumor and anti-tumor effects [52–55]. Schematically, the γδ T cells producing IL-17 represent the main subset with pro-tumor effects, while γδ T cells have an antitumor effect through tumor cell elimination [56–58]. This last effect is mediated by the release of perforins and granzymes, surface and soluble death receptors such as Fas and tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL), and triggering of antibody-dependent cellular cytotoxicity (ADCC) [30, 39, 41, 59]. IL-17-producing γδ T cells are usually Vδ1, whereas killing properties are shared by both Vδ1 and Vδ2 T cells [30, 39, 41, 59]. However, this skewed subdivision is instrumental in trying to understand the complex interactions between the TME and the immune system. IL-17 is a pleiotropic cytokine that has pro-inflammatory effects linking the innate and the adaptive immune response, favoring the recruitment of neutrophils, and its dysregulation can provoke some autoimmune diseases [57]. In principle, the triggering of autoimmunity should be considered as a positive event in the context of a tumor [60, 61]. Indeed, the anti-tumor therapy with the so-called immune checkpoint blockers (ICB) can induce the exacerbation or appearance of autoimmune reactivity and diseases as an undesired side effect in association with an optimal antitumor response [62, 63]. It has been demonstrated that Vδ1 T cells can recruit myeloid cells in the tumor [58]. These cells can differentiate in the TME as myeloid-derived suppressor cells (MDSCs), inhibiting the activation of CD8⁺ T cells through many inhibitory mechanisms, such as arginase 1 (ARG-1) and indoleamine 2,3-dioxygenase (IDO) [64, 65]. Notably, IDO expression is triggered by pro-inflammatory cytokines such as IFN-γ [66, 67]. On the other hand, IFN-γ can increase the expression in tumor cells and tumor-associated fibroblasts/mesenchymal cells of the programmed cell death protein ligand 1 (PD-L1) ligand of programmed cell death protein 1 (PD-1) [68–70]. The interaction between the PD-1 on cytotoxic T lymphocytes (CTLs) and PD-L1 present on the TME leads to further inhibition of CD8⁺ T cell antitumor activity. In conclusion, the IL-17 produced by Vδ1 T cells can strongly impair the adaptive response against tumors [58].

Although reported in breast adenocarcinoma and prostate cancer, instead of CRC, tumor-infiltrating Vδ1 T cells exerted a potent suppressor function on DC maturation and function controlled by the TLR8 signaling pathway. Indeed, the use of Poly-G3 and ssRNA40 as ligands for TLR8 expressed on Vδ1 reversed their suppressive functions. These findings suggest that innate ligands can shape the suppression activity of Vδ1, further complicating the cross-talk among the different components of the TME [71].

Rodin and coworkers [51] used mass and flow cytometry, mRNA quantification, and TCR sequencing to show that the composition of γδ T cells in the macroscopically healthy and CRC mucosa is heterogeneous among patients, and it is composed of Vδ1, Vδ2, and non-Vδ1-Vδ2 γδ T cells. This double-negative cell population expresses mainly the Vδ3 gene and minority Vδ4, Vδ5, Vδ7, and Vδ8 TCR genes [51]. Vδ2 T cells paired with Vγ9 mainly, although some Vδ2 T cells were Vγ9 negative. Overall, this analysis of TCR repertoire indicates a wide variety of γδ T cells in the CRC among the diverse patients. The key feature in CRC, compared to healthy colon mucosa, is the detection of an increase in non-Vδ1-Vδ2 T cells expressing Vδ3. However, in the relatively small group of patients analyzed (n = 49), there was no statistically significant correlation between the proportion of non-Vδ1-Vδ2 cells and microsatellite status, grade of differentiation, stage, location (right vs. left-sided), or patient age.

The mass spectrometry analysis indicated that peripheral blood and CRC mucosa γδ T cells formed distinct clusters. Furthermore, Vδ1 and Vδ2 cells express markers associated with cytotoxic effector functions. On the other hand, the non-Vδ1-Vδ2 cells appeared to express genes for proteins with tumor-promoting functions, such as neutrophil-recruiting chemokines, Galectin 3, and transforming growth factor-beta [51]. Importantly, the functional analysis of non-Vδ1-Vδ2 cells (mainly Vδ3) indicated that these cells express lower levels of IFN-γ and TNF than Vδ2 cells. Furthermore, IL-17 was expressed in non-Vδ1-Vδ2 cells in line with what was reported by other authors [72] and in murine systems [54, 73]. Altogether, these findings suggest that subsets of γδ T cells may play different roles in the CRC TME. It should also be noted that the number of γδ T cells present in the CRC mucosa is limited compared to other T cells, such as CD4⁺ Th17 [74]. This finding would indicate that CD4⁺ Th17 may have a preeminent role in pro-tumor effects compared to γδ T cells expressing IL-17. Importantly, the intraepithelial, but not stromal, localization of IL-17-producing CD4⁺ T cells positively correlated with improved survival [74]. Furthermore, IL-17 may trigger CRC angiogenesis [75]. On the other hand, IL-17 may recruit cytotoxic CD8⁺ T lymphocytes into the tumor [74, 76]. These findings indicate that it is not correct to assign a specific pro- or anti-tumor activity to IL-17. IL-17-producing cells may show different effects on the growth of CRC tumors depending on their localization, frequency, and intrinsic features.

It has been reported in CRC that high levels of the TRDV1 gene were not associated with a positive prognosis. Furthermore, gene expression correlation indicated that γδ T cells were more closely related to NK and innate lymphoid cells (ILCs) than CD4⁺ and CD8⁺ T cells. The Vδ1 T cells expressed low levels of some inhibitory receptors, such as PD-1, LAG-3, and T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), but were positive for T cell immunoreceptor with Ig and ITIM domains (TIGIT) [72]. Importantly, the Vδ1 T cells in CRC can produce IFN-γ or amphiregulin (AREG). AREG is an EGF-family ligand that plays a key role in epithelial cell proliferation, differentiation, and tissue regeneration via the epidermal growth factor receptor (EGFR)/mitogen activated protein kinase (MAPK) signaling pathway. This factor was found to be associated with microsatellite stability (MSS) compared to microsatellite instability (MSI) and the consensus molecular subtype 2 (CMS2) of CRC or a wound-healing phenotype of CRC. It is of note that γδ T cells were the only leukocyte population increasing the expression of AREG in CRC compared to healthy tissue. In addition, the AREG produced by Vδ1 T cells can trigger the proliferation of the two CRC cell lines SW480 and HCT116, as well as migration in a wound healing assay. These effects were inhibited with the anti-EGFR antibody cetuximab (Cet), which in turn blocks the interaction of AREG with the EGFR. It is of note that some Vδ1 T cells present in CRC can produce both AREG and IFN-γ or IFN-γ only. This would suggest that Vδ1 in the TME can show pro- and/or anti-tumor features. It is of note that the use of IL-15 can downregulate the production of AREG, increasing that of IFN-γ by Vδ1 T cells, while IL-1β is a potent stimulator for AREG production. These findings indicated that the cultivation of tumor-infiltrating and peripheral blood Vδ1 T cells with some specific cytokines can skew the type of Vδ1 T cells. Indeed, the possible use of Vδ1 T cells for antitumor therapy should avoid the presence of Vδ1 T cells producing AREG and favoring the growth of the tumor, and this can be accomplished with cultivation in IL-15 [72]. These considerations may explain the heterogeneity of function of Vδ1 T cells in CRC and highlight the role of the cytokine microenvironment in determining the properties of infiltrating cells. In the cohort of CRC analyzed, the Vδ2 T cells were reduced in CRC mucosa compared to the healthy one, and some of them can produce AREG, although the majority produced IFN-γ and a minority both AREG and IFN-γ. This would suggest that Vδ2 T cells may produce AREG, and this property should be considered for adoptive immunotherapy.

However, not all reports are in line with the apparent role as suppressor/pro-tumor cells of Vδ1 T cells. Indeed, the expression of the activating receptor NKp46 identified the largest fraction of IEL subset [77]. Importantly, the gut tropism and the ontogeny of these cells are linked to the ontogeny and gut-tropism of NKp46⁺ Vδ1 IELs depends on gut-environmental factors and some specific features. Low frequencies of the NKp46⁺ Vδ1 T cells in healthy gut specimens of patients affected by CRC were associated with faster tumor progression and development of metastatic diseases. These contrasting results with the other reports above can be explained by the different technical approaches, reagents, and gene or protein level to analyze the IL-17⁺ T cells as well as the intra- and inter-donor heterogeneity (IDH) of CRC tumors.

Besides Vδ1 T cells, the Vδ2 T cells have been described as important anti-CRC effector cells associated with a better prognosis in MSI CRC [78, 79]. Indeed, using a deconvoluted CIBERSORT matrix to investigate the abundance of Vγ9Vδ2 T cells, it has been reported that CRC with a high infiltration of this cell subset had better survival than those with a low infiltration. Importantly, it appears that a better overall survival (OS) was also associated with αβ T cells. In this case, a clear enrichment in gene sets for “response to IFN-α”, “inflammatory response”, “antigen processing and presentation”, “T cell activation”, and “cytolytic activity” was found with αβ T cells but not with γδ T cells. These findings suggest that the infiltration with either αβ or γδ T cells can always lead to a better anti-CRC response, but in a different immunological context.

Looking at what is described in this review regarding Vδ1, Vδ2, non-Vδ1-Vδ2, and Vδ3 lymphocytes, it appears that some subsets of CRC mucosa γδ T cells may have protumor and antitumor effects (Figure 2).

The pro-tumor effects are associated with IL-17 expression in mucosal Vδ1/Vδ3 T cells, while anti-tumor effects are associated with Vδ2 T cells. This clear-cut subdivision is challenged by what has been recently reported by Ran and colleagues [49]. These authors analyzed nine published gene datasets of single-cell RNA from 18,483 γδ T cells out of 951,785 total cells from the neoplastic or adjacent healthy tissue of 165 human CRCs. The bioinformatics analysis revealed that γδ T cells overall exhibited a cytotoxic-related transcriptional pattern in both CRC and healthy mucosa. Unexpectedly, none of the γδ T cell clusters showed an IL-17-related pattern, including master transcription factor RAR-related orphan receptor gamma (RORC), IL-23R, CCR6, and MAF, suggesting the inability to produce IL-17. Different subsets of γδ T cells could be defined as effector, tissue-resident memory, progenitor exhausted, and exhausted cells, and the authors noted an increased expression of cytotoxic molecules [49]. To identify γδ T cells from unsorted whole tissue, the combination of CD3D, CD3E, CD3G, CD247, and TRDC was considered as the minimal set of markers to be applied [49]. Overall, these findings propose that datasets from mice or humans may differ in their typical features for identifying specific T cell subsets or that the purity of selected lymphocyte subsets may be insufficient for single-cell analysis, potentially yielding misleading results [49, 80].

Notably, γδ T cells infiltrating different types of tumors do not show the exhausted functional behavior typical of CD8⁺ TCR αβ T cells, although they express several immune checkpoint molecules [81, 82]. Furthermore, it is to be noted that there are many possible explanations for these inconsistent findings. Firstly, the methods of detecting IL-17-producing cells may have different sensitivities, as well as detecting a different step for showing the relevance of IL-17. Typically, single-cell RNA sequencing (sc-RNA-seq) can detect tiny cell subsets comprised in the whole cell population analyzed, as in the case of γδ T cells. The mRNA presence does not mean the protein presence, and on the other hand, the antibodies used in IHC and flow cytometry (FC) can lack the good avidity and affinity for the low amount of IL-17 present in the CRC TME. The isolation procedures to analyze either sc-mRNA or IHC/FC can select some subsets of a given population, thus leading to misinterpretation of data. Finally, the huge amount of data coming from omics analyses needs well-established and controlled pipelines. To avoid any misleading message, it would be necessary to validate each other’s data obtained with advanced molecular or more classical protein-based approaches.

Table 1 is a summary of the main features of γδ T cells in healthy colon and CRC both in mice and humans, to point out the experimental model and the major differences among γδ T cells in these two species.

The frequency of γδ T cells is usually low in the PBMC, ranging from 1–10% roughly. The two main subsets in humans are represented by Vδ2 (the majority) and Vδ1 T cells (the minority). To generate a good enough number of γδ T cells from PBMC, it is necessary to expand them using appropriate culture conditions [94, 95]. Ideally, the use of an antigen could be a better means to achieve this aim. It is unclear whether γδ T cells can respond to specific polymorphic antigens; at least they do not respond to classical antigen peptides presented by antigen presenting cells (APCs) as αβ T cells [30, 41, 96, 97]. The γδ T cells recognize monomorphic molecules that function as the superantigens for some groups of αβ T cells [98, 99]. Among the molecules recognized by the TCR of γδ T cells are microbial metabolites such as HMB-PP, isoprenoid precursors such as IPP or dimethylallyl pyrophosphate (DMAPP), Hsps (Hsp60 in humans and Hsp65 in mice, Hsp70, and Hsp90), and lipid antigens presented in the context of CD1d molecules [100–109] and other specific molecules among which EphA2 [97, 110–117] (Table 2).

It is well known that human Vδ2⁺ T cells can be expanded using IPP or DMAPP. These phosphate antigens (pAg) are intermediates of the cholesterol synthesis that can be produced in epithelial cells, monocytes, and mesenchymal stromal cells (MSC) [118–120]. These pAgs can trigger both the activation and consequent proliferation of Vδ2⁺ T cells as well as the triggering of cytolysis directed to the pAg-producing cells or the production and release of pro-inflammatory cytokines such as IFN-γ and TNF-α [121, 122]. For presentation, some molecules belonging to the butyrophilin (BTN) family are necessary [123–126]. In particular, it has been demonstrated that BTN3A1 and BTN2A1 are the cell surface ligands presenting pAgs undergoing structural rearrangement. Specifically, BTN2A1 binds the Vγ9 domain, and BTN3A1 binds the Vδ2 and Vγ9 chains of the Vγ9Vδ2 TCR, rapidly triggering activation [123–126].

Conceivably, the stimulation of Vγ9Vδ2 T cells with pAgs will take place when BTN2A1 and BTN3A1 are expressed [120, 127–129]. Thus, the two components of the stimulatory armamentarium should be present: pAgs and BTN [124]. Practically, Vγ9Vδ2 T cells can be expanded from unseparated PBMC by adding IPP to the culture medium and exogenous IL-2 [118, 130]. This system, which needs the use of a purified chemical compound such as IPP, can be easily substituted by culture with micromolar concentrations of aminobisphosphonates (N-BPs) such as zoledronic acid (ZA) or risedronic acid (RIS) [131–134]. The N-BPs can interfere with cholesterol biosynthesis, inhibiting the enzyme farnesyl pyrophosphate synthase (FPPS) in the mevalonate pathway, preventing the prenylation of small GTPases. The effects of N-BPs are mainly of two types: an increase in the intracellular content of IPP and DMAPP, together with interference with the cytoskeleton arrangements by affecting the activation of Rho, Rac, and Cdc42 proteins [135–137]. Starting from unseparated PBMCs with a low percentage of Vγ9Vδ2 at the onset of the culture, the use of ZA and the addition of low doses of IL-2 (30 IU/mL) can induce a strong proliferation of Vγ9Vδ2. This can reach 60–90% of the culture after 10–12 days and more than 95% on day 14. This can be an easy method to obtain enough Vγ9Vδ2 for making immunotherapy. The expansion of Vγ9Vδ2 can be triggered using other cytokine cocktails containing IL-15 and other stimuli [138, 139]. In principle, TCR triggering by pAgs can be considered an easy way to generate large numbers of Vγ9Vδ2 T cells for immunotherapy. Importantly, the type of Vγ9Vδ2 T cell expanded cells is mainly composed of effector memory (EM) CD45RA^–CD27^– and central memory (CM) CD45RA^–CD27⁺/CCR7⁺ cells, while naive (N) and terminally differentiated effector memory (TEMRA) cells CD45RA⁺CD27^–/CCR7^– represent a minority [95, 140, 141]. These cells express functional activating receptors such as Natural Killer Group 2, member D (NKG2D, CD314), and DNAX Accessory Molecule (DNAM)-1 involved in the killing of tumor target cells in different cancers [142–145]. The ligands of NKG2D are the major histocompatibility complex class I (MIC)-related molecules A and B, as well as UL-16 binding proteins 1-6 (ULBP1-6) [146–148]. These molecules are upregulated on tumor cells, and the NKG2D-NKG2DL interaction can deliver signals leading to proliferation, cytotoxic granule exocytosis, and cytokine production [146–148]. Also, NKG2DL can be shed through the action of a disintegrin and metalloproteinase (ADAM) 10 or MMP downregulating the immune function [148]. Furthermore, DNAM-1 is physically and functionally linked to lymphocyte function associated antigen 1 (LFA1) adhesion molecule [149]. It can interact with CD155 (poliovirus receptor) and CD112 (Nectin-2) expressed on stressed tumor cells, leading to similar effects mediated by NKG2D engagement [150]. Importantly, CD112 can interact with poliovirus receptor related immunoglobulin domain containing (PVRIG/CD112R) and TIGIT, while CD155 can interact with TIGIT and T cell activation, increased late expression (TACTILE/CD96). TIGIT, TACTILE, and PVRIG are potent inhibitory receptors expressed on effector lymphocytes, and consequently their engagement can strongly inhibit the signal mediated by activating receptors such as TCR, NKG2D, and DNAM-1 [151–154]. Overall, the final results after the interaction between γδ T cells [88, 89, 91, 92] and tumor cells may depend on the balance between the positive and negative signals delivered into effector cells such as NK cells [154] (Figure 3).

It is to be noted that Vγ9Vδ2 shows a very low level of expression of PD-1 with a peak at days 3 and 4, with a consequent decrease to baseline levels after 7 days [155]. On day 14, the large majority of Vγ9Vδ2 T cells are negative for PD-1 expression [156]. The upregulation of PD-1 and TIM-3 appeared to be related to the dose of ZA used, but this inhibitory molecule was not induced using the mycobacterium Bacillus Calmette-Guérin (BCG) [157]. Importantly, the cytotoxic activity of Vγ9Vδ2 T cells expanded with 10 µM ZA was impaired, but not that of those expanded with low doses such as 0.5–1 µM ZA [157]. In our hands, the optimal stimulation and expansion of Vγ9Vδ2 T cells can be achieved only with low doses of ZA (0.5–1 µM), while higher doses (more than 5 µM) can result in reduced expansion of Vγ9Vδ2 T cells [158]. These findings suggest that the optimal dose to trigger proliferation of Vγ9Vδ2 T cells is quite low and is associated with very low, transient PD-1 expression. This supports the idea that the expansion with pAgs is a good choice to plan adoptive immunotherapy [159–163]. As already mentioned, γδ T cells express HLA-I receptors involved in the regulation of activation and tumor cell killing [164–166]. While the CLIR type of inhibitory receptors NKG2A/CD94 complex is expressed on a large fraction of γδ T cells [89, 164–167], the expression of KIR is more discrete on peripheral blood γδ T cells [88, 167, 168]. Importantly, the functional role of KIR on γδ T cells is controversial [10, 93]. The presence of inhibitory receptors such as KIR should deliver a negative signal blocking activation [9, 10, 88, 91, 93]. However, it is evident that Vδ1 and Vδ2 T cells can express isoforms of KIR that do not deliver an inhibitory signal into γδ T cells but can trigger activation [93]; likewise, it can happen in NK cells [169–171]. In detail, we have shown that γδ T cells can express KIR2DS1 and KIR2DS2 based on reactivity with monoclonal antibodies (mAbs) that recognize the extracellular portion of both inhibitory and activating discrete HLA-I alleles. The presence of the activating isoform is functionally demonstrated by the activation of target cell killing of the mouse FcγR (such as the murine mastocytoma P815, Figure 4) using the specific murine anti-human KIR mAb [93].

It is of note that the γδ T cell clones (either Vδ1 or Vδ2) expressing KIR2DS1 or KIR2DS2 can interact with the soluble HLA-I molecule alleles Cw4 or Cw3, respectively. This interaction with their natural ligands can deliver an apoptotic signal leading to the death of γδ T cells [93]. Overall, these findings would indicate that the presence of activating KIR on γδ T cells can regulate their survival [93]. In addition, sizeable fractions of γδ T cells express the CD8 receptor for the invariant portion of HLA-I [93]. Similarly to activating KIR, this receptor engaged with soluble HLA-I can trigger apoptosis of γδ T cells through the Fas-L-Fas interaction [93]. Altogether, these findings suggest that γδ T cells bear several HLA-I receptors that can regulate their activation and survival [93]. There is no direct evidence that these receptors can play a role in the context of CRC. It is possible to speculate that within the TME, the soluble (s) or surface-expressed HLA-I on CRC cells, MSCs, or regulatory monocytic cells may regulate the γδ T cell antitumor response. Theoretically, the function of sHLA-I, possibly derived from CRC cell killing, interacting with either activating or inhibitory receptors on γδ T cells may influence, depending on the rate of positive and negative signals delivered, their anti-tumor effect by interfering with the direct binding to CRC cells. This effect could be present with MSCs and other stromal cells such as monocyte-derived suppressors or dendritic cells. It is still undefined whether the amount of sHLA-I may indeed deliver a signal in the CRC TME. However, the lower physical interaction with tumor cells and TME stromal cells can reduce the signals delivered upon the engagement of activating or inhibitory receptors expressed on γδ T cells, modifying the outcome of γδ T cell-mediated activities. Finally, the results of the whole signaling mediated by either sHLA or surface HLA-I may be heterogeneous in different sites of the CRC mass. This heterogeneity can lead to the associated production of different cytokines, which in turn may influence the behavior of the TME. These hypotheses further complicate a very intricate scenario. The idea of blocking the activity of γδ T cells expressing receptors for HLA-I can be a solution to reduce the undesired effects of apoptosis mediated by CD8 and/or KIR2DS1/S2 and other activatory HLA-I receptors. This can be accomplished using γδ T cells lacking these receptors for immunotherapy or by blocking the HLA-I-γδ TCRs with monovalent anti-KIR2DS1/2 antibody-derived fragments to limit the apoptotic signal delivered by the receptor crosslinking with HLA-I [93]. However, all these considerations need an experimental demonstration to support tools targeting molecules that can influence at the same time the antigen-specific activity of αβ T cells.

Some phenotypic and functional features of γδ T cells are different in murine models from what happens in humans [8, 172–176]. Overall, the role of Vδ1 T cells is similar in mice and humans, while it is difficult to mimic in a mouse model the role of Vδ2 T cells, as this cell subset is almost absent in mice [174]. Furthermore, the human Vδ2 T cells, compared to Vδ1 T cells, can be more easily expanded and used as effectors in vitro, and they can be considered the more suitable subset of γδ T cells for immunotherapy. The functional features of Vδ2 T cells during the interaction between CRC epithelial cells can be studied in the three-dimensional (3D) culture systems such as patient-derived tumor spheroids and organoids [120, 129, 134, 158, 177–183] (Figure 3). These two 3D culture systems show some important differences to consider [184–187]. Spheroids are usually composed by the aggregation of a single cell type in a few days, and they generally do not need the presence of extracellular matrix to grow. On the other hand, organoids are more complex 3D structures forming crypts and a lumen and require specific matrix proteins for growth, besides several peculiar medium components. The organoids are composed of different cell types coming from stem cells differentiating into epithelial cells lining intestinal crypts, enterochromaffin neuroendocrine cells, and goblet cells. Intuitively, the growth of an organoid is slow and typical for each donor if they are derived from CRC mucosa specimens, and it is more difficult to standardize the culture conditions of organoids than those of spheroids. Importantly, both CRC spheroids from established cell lines and CRC organoids from several patients can be used as targets for γδ T cells. These models are ideal to study the molecular mechanisms of target cell recognition and the means of killing exerted by cytotoxic effector cells. In addition, these 3D structures can be used to select the drugs that are efficient both in single and combo administration and to analyze the mechanisms of insurgence of resistant cells [184–187]. It is of note that these analyses need advanced microscopy tools and appropriate bioinformatic analysis of many images to study these interactions. The inter-tumor heterogeneity (ITH) is related to the large variety in phenotype and treatment response of each patient. While the IDH is due to the cell products derived from PBMC of different donors. These two heterogeneities have been studied using “phenoscaping” during the interaction between CRC organoids and γδ T cells [177]. This advanced experimental approach uses high-dimensional single-cell analysis employing sc-RNA-seq and/or cytometry to determine how perturbation of a cellular system leads toward a specific state.

Applying thiol-reactive organoid barcoding in situ mass cytometry [188–192], it has been studied the posttranslational modification signaling, cell activation state, apoptosis, and immunophenotype of γδ T cells interacting with patient-derived organoids (PDO) at the single-cell level resolution [177]. Overall, the analysis of more than 1,000 γδ T cells in these 3D culture systems showed that tumor cells can suppress γδ T cell-mediated cytotoxicity, but γδ T cells can overcome this immune suppression by using multimodal tumor cell killing. Indeed, phenoscaping indicated that unmodified γδ T cells showed poor cytotoxicity against CRC PDO. On the other hand, γδ T cells expressing an IL-15Rα–IL-15 fusion protein can efficiently kill PDO, but this cytotoxicity can be downmodulated by specific PDO, indicating an ITH-specific immune modulation. This effect was overcome through anti-B7-H3 mAb obtained from a murine single-chain variable fragment (scFv) phage-display library [177]. This was associated with an hIgG1 format using a G1m1, 17 heavy chain allotype and produced as a full antibody able to trigger ADCC. Furthermore, the engineered γδ T cells can kill PDO enriched in cancer stem cells refractory to chemotherapy. Overall, the ADCC can overcome the immunomodulation due to PDO, and it suggests that γδ T cells can kill PDO CRC targets through different molecular mechanisms. This notion is further supported by several experimental evidence obtained using co-cultures of γδ T cells and PDO from CRC mucosa [129]. In fact, it has been demonstrated that CRC PDO can trigger the expansion of Vδ2 T cells when exposed to 5 µM ZA starting from highly purified peripheral blood T lymphocytes. Furthermore, this effect was evident using the antibody-drug conjugate (ADC) composed of the anti-EGFR therapeutic antibody Cet conjugated with ZA (Cet-ZA) [129]. The Vδ2 T cells obtained in the co-cultures with PDO killed PDO cells in the absence of any additional stimulus at quite a high effector-target cell ratio (E:T ratio 20:1), while at a low E:T ratio Vδ2 T cells did not kill PDO cells spontaneously. However, at a low E:T ratio, the addition to the cell cultures of either soluble ZA, Cet, or Cet-ZA can trigger the CD107a degranulation and killing of tumor cells. Importantly, the effect of the Cet-ZA can be partly inhibited through anti-TCR mAb specific for the Vδ2 chain and anti-CD16 antibody. These findings strongly suggest that the Vδ2 T cells can use both TCR and FcγR to target CRC PDO, supporting again the multimodal ability of Vδ2 T cells to kill tumor cells [129]. Importantly, the single-cell analysis has allowed the identification in a murine model of the key role of IL-17-producing γδ T cells [178]. This was associated with the analysis of organoids at different stages of evolution of damage at anorectal junctions representing hyperplasia, dysplasia, and carcinoma stages [178].

On the other hand, the use of spheroids from well-established CRC cell lines can provide important information on the functional features of tumor cells in 3D masses. Indeed, we have shown that CRC cell lines can generate spheroids with peculiar physical features by evaluating the spheroid mass density. This mass density was relevant during the killing of tumor cells exerted by NK cells [193].

The killing of CRC tumor cells induced upon the interaction with γδ T cells was mediated by both the engagement of TCR and ADCC. Furthermore, the CRC spheroids and organoids can be used as a target of nanoparticles carrying N-BPs to study novel putative therapeutic tools to trigger γδ T cells’ anti-tumor activity [181]. Figure 3 shows some of the features and applications of spheroids and organoids.