Molecular and cellular players within the thymic microenvironment are crucial for acquisition of γδT cell effector fate (Figure 1). Anatomically, precursors migrate from the cortex to the medulla during this process, where they receive multiple signals that dictate their differentiation. In this three-dimensional environment, it is important to keep in mind that time should be considered as a fourth dimension when deciphering the “natural” γδT17 cells ontogeny.

Many cytokines have been reported to directly or indirectly regulate thymic γδT17 cell differentiation/development.

IL-7 is a critical and non-redundant cytokine in lymphopoiesis (62). IL-7Rα-deficient mice completely lack γδT cells (63, 64) partly due to the role of IL-7 in V-J recombination of TCRγ genes (63, 65). Its specific role in survival and proliferation of γδT17 cells in peripheral tissues under homeostatic and pathological situations has been demonstrated (37, 66, 67). Even if this question has never been directly addressed, many studies imply a requirement for IL-7 in the proper natural γδT17 cell development. First, in a model of conditional abrogation (RBP-Jκ) that precludes IL-7Rα (CD127) expression, but in which the initial generation of γδ precursors (DN2 and DN3 stages) is maintained, the pool of both thymic and peripheral γδT17 cells is markedly reduced (68). Second, addition of recombinant IL-7 in fetal thymic organ culture from E16 thymus promotes γδT17 cell expansion over IFN-γ-producing γδT cells (37). Understanding whether IL-7 directly participates in the differentiation program or solely in the post-differentiation expansion of γδT17 cells will require further investigations. In line with the favorable temporal window for γδT17 cell development around birth, it is important to mention that thymus is particularly enriched for IL-7 [especially in thymic epithelial cells (TECs)] in neonates and its presence declines with age (69, 70).

Akin to conventional αβ Th17 cells, TGF-β1 has also been proposed to participate in γδT17 cell development (71). In addition, this report indicates that some other Th17-driving cytokines, such as IL-23 and IL-21, appear to be dispensable. However, since the authors have performed their investigation in 11-day-old mice, a time at which natural γδT17 cells already egressed the thymus, it is difficult to precisely evaluate the contribution of TGF-β1 in natural γδT17 cell differentiation.

The role of IL-6 in γδT17 cell development is somehow controversial. While several studies indicated that IL-6 deficiency did not influence γδT17 cell homeostasis and cytokine production capacity (71, 72), others reported a defect in γδT17 cells in both thymus (73) and peripheral organs (74). Thus, the contribution of IL-6 in natural γδT17 cell differentiation remains unclear. In addition, mRNAs for IL-6Rα (CD126) are barely expressed on both immature and mature thymic natural γδT17 cells (from the ImmGen database). However, it cannot be firmly excluded that IL-6 influences the thymic microenvironment to support γδT17 cell differentiation as previously hypothesized (73).

Thus, this indicates that the cytokine network within the thymic microenvironment has to be tightly regulated in a time-dependent manner to allow natural γδT17 cell differentiation.

Akin to other thymocytes, interaction with TEC is likely critical in γδT17 cell differentiation, even though few experimental data are currently available. Thus, medullary (m)TEC has been implicated in regulating Vγ6⁺ γδT17 cell development through the TF autoimmune regulator (Aire) (69). In Aire^−/− mice, IL-7 production is up-regulated in mTEC, and this is accompanied by an overproduction of Vγ6⁺ γδT17 thymocytes. Interestingly, other subsets of natural γδT17 cells, especially Vγ4⁺ subsets were not affected by specific Aire deletion in mTEC (69). This feature also indicates that the various natural γδT17 cell subsets probably require different signals to develop.

Cortical (c)TEC has also been recently shown to control γδT17 cell development. Using a mouse model with specific cTEC ablation, Nitta and colleagues observed a strong dysregulation in the proportion of natural γδT17 cell subsets (75). Specifically, absence of cTEC skewed the γδT17 TCR repertoire toward Vγ6 expression at the expense of the Vγ4⁺ T cell subset during the postnatal period while the proportions of Vγ6⁺ and Vγ4⁺ γδT17 cells remained normal during embryonic life. Authors hypothesize that in their mouse model, the postnatal thymic microenvironment resembles to the fetal microenvironment, which in turn favors the Vγ6⁺ subset. In addition, since they have been proposed to be a prime source of TGFβ (76), cTEC could also participate in thymic development of γδT17 cells through this mechanism (71).

The importance of Dll4, a Notch ligand expressed by TEC (77) has also been proposed in γδT17 cell differentiation (78). In a co-culture model of E15 thymocytes with stromal cells, absence of Dll4 expression by stromal cells led to an abrogation in γδT17 cell development (78). This phenotype indicates that Dll4 is likely to be a common factor for the differentiation of all natural γδT17 cell subsets.

Finally, a recent study proposed that signaling through NF-κB-inducing kinase (NIK) in TEC is essential for the generation of a fully functional pool of γδT17 cells (79). However, the molecular factors regulated by NIK in TEC are yet to be determined. Understanding the NIK-dependent pathways in TEC will certainly provide important clues in the TEC-γδT17 precursor interaction mechanisms that drive γδT17 cell effector fate.

Altogether, the fetal and perinatal thymic environment offers a temporal window of opportunity for γδT17 cell differentiation. However, the available literature indicates the requirement for differential factors according to the subset of naturally occurring γδT17 cells. This might somewhat rely on the intrinsic nature of the γδT17 precursors. It can also be hypothesized that these precursors (Vγ6⁺ and Vγ4⁺) require timely expressed TCR ligands in the thymic environment. However, no host-derived Ags have been proposed to date to participate in γδT17 cell differentiation.

On top of the TCR, its accessory receptors have been proposed to participate in γδT cell differentiation. In addition to be a convenient marker to distinguish γδT functional subsets, the costimulatory receptor CD27 has been shown to participate in γδT cell development (14). Indeed, γδ thymocytes from Cd27^−/− mice presented altered expression of ifng. Although CD27 deficiency did not influence the pool of γδT17 cells, CD27 gain of function in thymic cultures resulted in lower IL-17 transcripts by CD27⁻ γδ thymocytes (14). Thus, CD27 appears as a thymic regulator in γδT cell effector fate.

Inducible T cell co-stimulator (ICOS) signaling pathway has also recently emerged as a possible determinant in γδT17 cell (at least for the Vγ4⁺ subset) development (86). Agonistic activity of anti-ICOS mAb in fetal thymic organ culture significantly impaired Vγ4⁺ γδT17 development. In line, genetic ablation of ICOS tends to increase the pool of thymic Vγ4⁺ γδT17 cells (86). Thus, this study indicates that ICOS-dependent intracellular pathways in thymocytes controls γδT17 cell effector fate.

Besides, one must keep in mind, that, alongside with TCR and costimulation receptor signaling, multiple other signals have to be integrated by embryonic thymocytes to, in fine, engage toward the γδT17 effector fate.

γδT17 precursors express specific receptors for various soluble factors produced in the thymic environment by hematopoietic and non-hematopoietic cells. Signals provided by these mediators have to be integrated by γδT cells to fully develop.

For instance, TGFβ receptor (TGFβR) signaling pathway in developing γδT17 cells could be important in their effector fate. Mice deficient for Smad3, a critical component of the TGFβR signaling pathway presented a striking defect in frequency of thymic γδT17 cells compared with littermate controls (71). However, since the authors did not provide direct evidence (bone marrow chimera and OP-9 models) for an intrinsic role of the TGFβR signaling pathway, it cannot be excluded that this pathway is indirectly linked to γδT17 cell development.

In addition, the targeting of lymphotoxin-β receptor through double positive thymocytes-derived ligands (87) also controls the generation of γδT17 cells by regulating the expression of TFs from the NF-κB family namely RelA and RelB (88).

As mentioned above, the IL-7/IL-7Rα axis is important to generate a normal pool of γδT17 cells. However, a better understanding of the downstream molecular cascade involved will be helpful to understand whether this axis controls the differentiation program or the homeostasis of γδT17 cells.

On the other hand, IL-15Rα signaling disruption favors the development of γδT17 cells in thymus of neonates (65). The molecular mechanisms responsible for this are currently unknown but might rely on the activation of repressing factors in the IL-15Rα signaling pathway or could be indirect by reducing the competition with γδT17-driving γ-chain-dependent cytokines such as IL-7.

The expression of the prostacyclin (PGI2) receptor (IP) on thymocytes has also been demonstrated to control γδT17 cell development. This was evidenced by a failure to generate γδT17 cells in the thymus of IP^−/− mice (89). However, the molecular determinants involved in this process are yet to be defined.

The dynamic integration of these multiple signals leads to the implementation of a complex transcriptional program that dictates γδT17 cell effector fate (Figure 2). The aim of this program is ultimately to maintain and/or to favor the expression of Rorc (encoding RORγt), the cardinal TF for IL-17-secreting cells (90) including γδT cells (72). The recent advances in RNA deep sequencing analysis allowed a better understanding of the transcriptional regulation involved in γδT17 development.

Thus, SRY-related HMG-box (SOX) 4 and SOX13, two members of the high mobility group box TF family constitute a central node of regulation in this program (50, 91, 92). SOX4/13 are paramount in the acquisition of the γδT17 effector fate by (1) directly controlling Rorc transcription, (2) possibly enhancing important γδT17 cell-driving pathways such as IL-7Rα signaling, and (3) possibly inhibiting Rorc-repressing TFs (31, 92). In this later mechanism, SOX13 was suggested to inhibit the Rorc-repressing activity of two downstream mediators of the Wnt/β-catenin signaling pathway namely lymphoid enhancer-binding factor 1 (Lef1) and transcription factor 1 (TCF1) (92). In this context, Tcf7 (encoding for TCF1) deficiency leads to an aberrant high proportion of γδT17 cells (92). At this stage, it is also important to mention that this pathway also regulates the development of the IFN-γ-producing Vγ1Vδ6.3⁺ and Vγ5⁺ cells (92). The mechanisms by which the TCF1–Lef1 axis counteracts the γδT17 transcriptional program are not fully understood. First, it is possible that TCF1 and Lef1 control Rorc expression through epigenetic (histone deacetylase) activity as suggested in conventional T cells (93). Alternatively, this axis could also indirectly repress Rorc expression by inhibiting the transcription of B lymphocyte kinase (Blk) (92), an important signal transducer in γδT17 cell development (94). However, how Blk controls RORγt expression is currently unknown. It is noteworthy that, using blk^−/− mice, Vγ6⁺ T cells were shown to be more Blk-dependent than Vγ4⁺ (94). As SOX13 was shown to regulate Blk expression and to control Vγ4⁺, but minimally Vγ6⁺, subset development (92), these results appear somewhat contradictory. However, Vγ4 and Vγ6 subsets develop at different temporal windows; therefore, they are likely to integrate different thymic signals. As a result, this might modulate the relative importance of a same regulatory axis, and eventually leading to different effects on their respective transcriptional program. Thus, regulatory network required for Vγ4⁺ ontogeny may be more SOX13-dependent than Vγ6⁺ subset. Regarding the differential contribution of the TCR signaling in the γδT17 effector fate of these populations, Blk may play a role at this stage. According to its regulatory activity on TCR signaling, Blk could act as a “rheostat” to fine-tune signals delivered by the thymic γδT cell ligands. Thus, Blk deficiency might affect more Vγ6⁺ ontogeny as TCR signaling has been proposed to control the development of these latter but not Vγ4⁺. Of note, Blk overexpression has been shown to enhance IL-7 responsiveness in B cells (95).

As stated earlier, the TCR signaling pathway strongly influences the acquisition of the γδT17 effector fate. Mechanistically, TCR engagement induces upregulation of proteins of the early growth response (Egr) family namely Egr2 and Egr3 (31, 81). These two TFs positively regulate the DNA-binding protein inhibitor Id3. Thus, Id3 impairs γδT17 cell differentiation through (1) inhibition of HeLa E-box binding protein (HEB)-dependent Sox4 and Sox13 expressions (96) and (2) inhibition of the Rorc promoter E47 (91). This scheme is also in line with the differential requirement for TCR strength in γδT17 cell effector fate of Vγ6⁺ and Vγ4⁺ further emphasizing the differences in the developmental programs of Vγ6⁺ vs Vγ4⁺ γδT17 cells.

In addition, the promyelocytic leukemia zinc finger (PLZF) protein is a key TF in the development of some innate and innate-like lymphocytes that dictates their acquisition of a Th-like effector program (97–99). PLZF was shown to control Vγ6⁺ differentiation into γδT17 cells (100). The molecular mechanisms that govern PLZF activity in Vγ6⁺ development are currently unknown and require further investigations. PLZF contribution in Vγ4⁺ γδT17 cell development has not been assessed yet. However, it is noteworthy that PLZF does not appear to be expressed in neonates Vγ4⁺ precluding a role of this TF for this particular subset (100).

Among the intracellular pathways that dictate the γδT17 effector fate, Notch signaling contributes to the generation of γδT17 cells through the helix-loop-helix protein Hes1 (78). Since it mainly exerts transcriptional repressing activities, it is possible that Hes1 acts in one of the pathways discussed above. Unrevealing the Hes1 interactome in developing γδT17 cells will be informative to get insight into the molecular factors that regulate this mechanism. An additional pathway by which the Notch signaling pathway could influence γδT17 cell development is by promoting IL-7Rα expression through the RBP-Jκ pathway (68). However, this pathway was only described in peripheral γδT17 cells of adult mice to control their homeostasis and self-renewal.

In addition to its role in proliferation/survival of γδT17 cells during development, the IL-7Rα signaling pathway is also likely to contribute to their transcriptional program. Indeed, IL-7/IL-7Rα signaling in fetal thymocytes was demonstrated to blunt both Lef1 and Tcf7 expression (101).

In silico analyses have also been fruitful to understand the transcriptional program of γδT17 cells. Using an algorithm that predicts important regulators across various lineages, the TF ETV5, along with SOX13 was proposed as a master regulator in Vγ4⁺ γδT17 cell differentiation (102). Conditional ablation of ETV5 in T cells confirmed the role of this TF in Vγ4⁺ γδT17 effector fate (102). Absence of ETV5 in developing Vγ4⁺ slightly reduced RORγt expression but severely impaired IL-17 secretion. This is reminiscent with the situation for Th17 cell differentiation in which ETV5 directly promotes il17a and il17f expressions but has no influence on Rorc (103). The role of ETV5 in Vγ6⁺ development remains to be determined; however, ETV5 is highly expressed on immature fetal Vγ6⁺ and strongly repressed upon maturation (http://www.immgen.org/databrowser/index.html).

Originally thought to be acquired by “neglect,” this literature underlines that the γδT17 effector fate is under the control of a very active process in which the SOX4/13 axis acts as a guardian for proper RORγt expression. In addition, the discrepancies in the phenotypes observed for Vγ4⁺ and Vγ6⁺ γδT17 cells imply different programs for natural γδT17 cell development. Thus, further transcriptomic analyses at single cell resolution are clearly required to better decipher the overlapping and/or specific developmental “trajectories” that drive the effector fate of these subsets.

Along with Nr4a1, other genes encoding for nuclear receptors such as receptors of vitamin A (Rarg) and D (Vdr), two vitamins reported to participate in ILC, γδT and NKT (25, 106) homeostasis and development are up-regulated during Vγ6⁺ maturation. In addition, genes encoding for vitamin transporters (Slc23a2 and Slc2a3) were also modulated in the Vγ6⁺ gene set. This could imply an important role for vitamins in γδT17 cell development. Somewhat related, expression of gpr183, a sensor of oxysterols was strongly up-regulated (3.6 fold) in developing Vγ6⁺ T cells. Interestingly, GPR183 has recently emerged as a critical player in the control of ILC3 homeostasis (107). Moreover, oxysterols are ligands for RORγt and drive Th17 cell differentiation (108). The influence of nutrient-derived metabolites on lymphocyte immunity including early development has recently gained considerable attention (2). The recent discovery of vitamin B2 metabolites as Ags for MAIT cells (109) will certainly reinforce the interest of immunologists for nutrient metabolism. The availability of vitamins in utero has also been shown to control the quality of the immune system in later life (110). According to these arguments and the temporal window of development for γδT17 cells, investigating the nutrient metabolism in γδT17 cell biology will be likely to provide new interesting data on the influence of maternal diet in shaping immunity.

In line with the emerging concept of neuroimmunology, numerous members of the “neuroactive ligand-receptor interaction” pathway were present in the Vγ6⁺ gene set including genes encoding for neuroactive substance receptors of neuropeptide (Gpr83), hormones (Sstr2, Rxfp1, and Calcrl), prostanoids (Ptger4, Ptgfrn, and Ptgir), leukotrienes (Cysltr2 and Ltb4r1), nucleotides (P2rx7, Adora2a, and Lpar4), and amino acids (Gabbr1, Gabbr2, and Gria3). Interestingly, the above-mentioned genes encoding for prostanoids, hormones, and nucleotide receptors are down-regulated during Vγ6⁺ T cell maturation, while those encoding for leukotrienes are up-regulated.

The relationship between neuromodulators and immune cell development is largely unexplored. However, there are evidence for an expression of neuromodulators and their associated-receptors in TEC and on thymocytes, respectively (111). Furthermore, ex vivo somatostatin (ligand for Sstr2) addition in FTOC increased thymocyte numbers and maturation. By contrast, both neuropeptide Y (ligand for Gpr83) and calcitonin (ligand for Calcrl) reduced thymocyte numbers (112). Last, Ptgir (encoding for the prostaglandin I2 receptor) has already been biologically validated to participate in natural γδT17 cell development (89). Regarding this, a broad analysis of the eicosanoid family in γδT17 cell development should be encouraged.

Among highly regulated genes, our analysis also retrieved genes related to circadian rhythm. Thus, we found that Nr1d1, Nr1d2, and Bhlhe40, which encodes for REV-ERBα, REV-ERBβ, and Dec1 proteins, respectively, were all up-regulated in mature Vγ6⁺ T cells. These proteins are critical repressors of Arntl (encoding for Bmal1), Npas2, and Clock, three master clock genes (113, 114). Of note, Arntl, Npas2, and Clock were also significantly regulated in developing Vγ6⁺ T cells. Recent literature has emphasized the importance of circadian rhythms in fine-tuning immune responses (113, 115). Interestingly, RORγt expression has been shown to be under circadian regulation through a REV-ERBα-dependent mechanism (116). Thus, these data may suggest a role for the circadian rhythm in Vγ6⁺ T cell biology. Of note, NFIL3 (E4BP4), a repressor of the circadian clock, is implicated in the differentiation/development of ILCs and NK cells (117–119). In addition, we can speculate that the regulation of these clock genes helps mature Vγ6⁺ T cells to integrate and to regulate circadian cues once in the periphery, as demonstrated for many other cellular actors of innate immunity (113).

Regarding the growing interest for understanding immunometabolic pathways implicated in leukocyte biology, we searched for genes involved in metabolic pathways. Our analysis revealed that many genes involved in the six major metabolic pathways specifically glycolysis (Aldh2, Ldhb, Acss1, and Acss2), tricarboxylic acid cycle (Idh1, Idh2, and Aco1), pentose phosphate pathway (Fbp1), fatty acid oxidation (Cpt1a, Nr4a3, Peci, Auh, Pex5, and Ivd), fatty acid synthesis (Fads2 and Slc45a3), and amino acid metabolism (Aco1 and Bcat1) are down-regulated in mature Vγ6⁺ T cells.

Interestingly, fatty acid oxidation is associated with regulatory T cell differentiation while glycolysis is a major metabolic pathway in effector T cell differentiation (120, 121). Of note, expression of Cpt1a was reduced in Th17 cells compared with regulatory T cells (120). Enhanced fatty acid synthesis and glycolysis in immune cells, especially T cells, have been regarded as markers of inflammatory cells required for acquisition of effector functions upon inflammatory conditions (122). This adds a metabolic argument into the fact that Vγ6⁺ T cells are “preset” cells with metabolic programing occurring during development, to be immediately and fully functional once in peripheral tissues.

To date, there is limited literature on the role of microRNAs in leukocyte development. In our gene set, we detected the presence of five microRNAs (mir15b, mir181a-1, mir181a-2, mir181b-1, and mir181b-2), all of them being down-regulated in mature Vγ6⁺ T cells. Of note, miR-181 was reported to be essential in NKT cell development (123, 124). Interestingly, the group of Immo Prinz studied the impact of miR-181a/b-1 deficiency on γδT cell development. While thymic Vγ1⁺ and Vγ4⁺ T cell subsets were unaltered in the absence of miR-181a/b-1, authors reported a higher frequency of thymic, but not peripheral Vγ6⁺ cells in miR-181a/b-1-deficient mice (125). The reason that underlines this feature is currently unknown. However, since miR-181 is a well-known positive regulator of the TCR signal strength (126), it is tempting to hypothesize that a reduced TCR signaling confers an advantage for Vγ6⁺ T cell differentiation. Thus, defining the miRome of developing γδT17 cells will potentially bring a novel layer of complexity in their developmental program.

Even if these in silico analyses suggest a role for scantily explored biological pathways in γδT17 cell development and/or maintenance, supportive experimental data are clearly required to explore these predictive hypotheses. It is noteworthy that these proposed biological pathways are not meant to be Vγ6⁺ cell-specific and, therefore, it does not preclude that other cell types including other γδT cell subsets may rely on similar biological pathways to develop. In addition, instruction that drives γδT17 effector fate is likely to start early in thymocyte development [even before TCR rearrangement (49)]; therefore, comparing immature vs mature populations probably induced an important bias in our analysis.