Disorderly conduct in γδ versus αβ T cell lineage commitment

doi:10.1016/j.smim.2010.04.003

Semin Immunol. Author manuscript; available in PMC 2011 August 1.

Published in final edited form as:

Semin Immunol. 2010 August; 22(4): 222–227.

Published online 2010 May 6. doi: 10.1016/j.smim.2010.04.003

PMCID: PMC2945703

NIHMSID: NIHMS204797

Disorderly conduct in γδ versus αβ T cell lineage commitment

Kavitha Narayan and Joonsoo Kang^*

Department of Pathology, Graduate Program in Immunology and Virology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655, USA.

^*To whom correspondence should be addressed: Joonsoo Kang, Department of Pathology, 55 Lake Avenue North, S3-137, Worcester, MA 01655, USA, Phone: (508) 856 2759, Fax: (508) 856 4336, Email: Joonsoo.Kang/at/umassmed.edu

The publisher's final edited version of this article is available at Semin Immunol

See other articles in PMC that cite the published article.

Publisher's Disclaimer

Abstract

The mechanism of T cell precursor commitment to the γδ or αβ T cell lineage remains unclear. While TCR signal strength has emerged as a key factor in lineage commitment based on TCR transgenic models, the entire TCR repertoire may not possess the same discriminatory power. A counterbalance to the TCR as the lineage determinant is the pre-existing heterogeneity in gene expression among precursors, which suggests that single precursors are unlikely to respond homogeneously to a given instructive signal.

Keywords: T cell development, T cell lineage commitment, T cell receptor, gammadelta T cells, WNT signaling

1. Introduction

Binary cell fate decision processes underpin the production of functionally distinct cell lineages of the mammalian immune system. For precursors committed to the T cell lineage, the earliest cell fate decision generates two types of T cells distinguished by the expression of either an αβ T cell receptor (TCR) or a γδ TCR on their cell surface. The resolution of the cell fate decision process has an indisputable order, leading to eventual commitment to either the γδ or αβ T cell lineage. However, whether the process itself is orderly, following a linear progression with defined checkpoints, or disorderly, with initial stochastic variations in gene expression and plasticity in precursors that attain lineage specificity as a function of a developmental clock, is poorly understood. γδ T cells are prototypic innate-like lymphocytes that share some phenotypic and functional similarities with NK and NKT cells. They are found predominantly in mucosal interfaces with the outside world, such as the skin, gut, and reproductive tract. How γδ T cells are generated in the thymus, and how they acquire the unique homing and functional capacities that set them apart from the conventional adaptive T cells expressing αβTCRs remains a challenging question.

2. Two models of lineage commitment

αβ and γδ T cells are derived from T cell lineage-committed double-negative 2 (DN2) precursors (CD3⁻CD4⁻CD8⁻CD25⁺CD44⁺c-kit⁺). TCR gene rearrangements are initiated during the DN2 to DN3 (CD3⁻CD4⁻CD8⁻CD25⁺CD44⁻) transition [1, 2] and productively rearranged Tcrg and Tcrd genes are necessary for γδ T cell differentiation, whereas successful Tcrb gene rearrangement alone is sufficient for DN3 cells to develop into immature αβ lineage CD4⁺CD8⁺ double positive (DP) thymocytes. Functional TCRβ chains pair with the invariant preTCRα (pTα) chain to form the preTCR complex. Over the years, two evolving models have occupied the field of γδ/αβ lineage commitment, centered on the possible role of the TCR as an on/off switch in the lineage decision process. The TCR instructive model proposes that the type of TCR, by virtue of different strengths of signaling, specifies the fate of bipotential precursor cells: relatively strong TCR signaling in DN2 and DN3 thymocytes specifies γδ T cell lineage fate, while signaling of lower intensity or duration wires precursors to become DP thymocytes [3, 4]. In contrast, the stochastic or TCR permissive/selective model proposes that a functional TCR simply permits the maturation of precursor cells that are already lineage-biased [5, 6]. The stochastic model also predicts that DN2 cells are not a uniform population that experiences quantitative variations in TCR signals, but rather a heterogeneous mixture of sub-populations marked by variations in gene expression that, once reinforced and fixed, become lineage-restricted [7]. In this scenario, while either preTCR or γδTCR expression is sufficient to permit the continued differentiation of DN precursors, αβ lineage precursors expressing a mismatched functional γδTCR will not mature since Tcrg gene transcription is silenced in DP thymocytes [8]. Conversely, γδ lineage precursors expressing a mismatched preTCR will not develop further since termination of Rag1/2 expression in TCR-signaled cells will prevent functional TCRα chain production [9–11], and the absence of a γδTCR or αβTCR will lead to the death of thymocytes.

Recently, direct probes of TCR signaling and of the molecular heterogeneity of early T cell precursors have yielded potential deterministic factors of T cell lineage commitment. This review will present a case against an instructive role for the TCR in γδ/αβ T cell lineage commitment and summarize evidence favoring the stochastic model.

3. Potential issues with the signal strength model

The recent conceptual framework of the instructive model consists of documented differences in signaling by the preTCR and γδTCR [12], and the precedent that other T cell lineage and subset differentiations are thought to be controlled by TCR signal strength or duration [13, 14]. Convincing experimental support for the quantitative instructive model of γδ/αβ T cell lineage commitment was provided by the laboratories of Love [3] and Wiest [4] using γδTCR transgenic (Tg) mice with altered signaling capacity. As detailed by others in this issue, in essence, decreasing the overall signal strength transmitted in T cells of two γδTCR Tg mouse lines consistently generated more CD4⁺CD8⁺ αβ lineage DP thymocytes, whereas increasing the TCR signaling capacity led to fewer DP cells (and sometimes more γδ T cells). Using KN6 γδTCR Tg mice, which recognize the β2m-dependent non-classical MHC class Ib molecule T22, it was shown that decreasing the signal strength through restriction of ligand availability resulted in the generation of αβ lineage DP cells [4]. However, in an alternate transgenic model of γδTCR interaction with T22 (G8) the ligand is not essential for the production of mature (CD24^lo) Tg⁺ thymocytes; rather it reduces the number of Tg⁺ thymocytes by negative selection [15]. Further, more recent analysis in non-transgenic mice has shown that γδ T cells reactive to T22 and T10 (which is highly homologous to T22) do arise in B2m^−/− mice, which lack the ligands, and they remain CD4⁻CD8⁻ with no down-regulation of the TCR, and show no skewing toward the αβ lineage [16]. Hence, these discordant results indicate the existence of clonal variations in the requirement for γδTCR signaling, and that the modulatory effects associated with select γδTCRs may not be experienced by most developing γδ cells. Finally, while these data show that signal strength can be an important parameter of cell lineage determination in TCR Tg mice, these studies still cannot distinguish between altering the initial lineage choice and confirming a lineage choice made previously through a stochastic mechanism.

A non-TCR Tg approach to assess the importance of TCR in T cell lineage commitment involves differentiation of precursors on OP9 stromal cells expressing the Notch ligand Delta-like-1 the (OP9-DL1) in which TCR signaling can be manipulated. On OP9-DL1 cells, single αβ lineage cells that have traversed through the β selection checkpoint by virtue of preTCR expression, known as DN3b cells, do not give rise to DP cells if signaled through the CD3 complex [17]. Instead, nearly all cells develop into CD4⁻CD8⁻ thymocytes, a portion of which express the αβTCR (or in a minority of cells, the γδTCR). Similar DN αβTCR⁺ thymocytes are especially prevalent in TCRαβ transgenic mice and have been shown to belong to the γδ T cell lineage [18, 19]. Hence, these in vitro studies suggested that lineage maturation is impacted by TCR signaling and that single preTCR⁺ DN3b cells that have been seeded and presumably weakly signaled by the preTCR (in some cases for as many as 8 days) are not lineage fixed. In mice and in fetal thymic organ cultures, the transit time between DN3 to DP thymocytes is relatively short (3 days or less, [20]). The reason for the extended duration for DN3b maturation in the culture system and the extent of gene expression variations in precursors arising from the delay are unknown. This uncertainty, combined with the heightened sensitivity of DP cells to TCR signaling in vitro [21], raises the possibility that DN TCR⁺ cell generation at the apparent expense of DP cells is not lineage conversion per se; but rather it can be the result of DP cell death during culture with anti-CD3 antibody, and consequent selective maintenance and expansion of rare DN TCR⁺ cells. Additionally, in reciprocal experiments where γδ T cells are generated in OP9-DL1 cultures by strong signaling (TCRδ crosslinking), the cells that develop are not “conventional” γδ T cells, but instead resemble NKT cells (γδNKT cells discussed in section 4). Also troubling, cell surface levels of CD5, which normally correlate with signal strength, were diminished rather than enhanced on cells that received TCRδ crosslinking. Finally, when transferred into mice, all of the OP9-DL1 culture-derived γδ T cells homed to the gut epithelium and few, if any, were found in the spleen and lymph nodes [17]. From these results it is not clear whether the cells generated by TCR crosslinking in this system are representative of normal T cell development in vivo.

Lastly, in all these transgenic and in vitro studies, a problematic and unexplained observation against the instructive model is that normal polyclonal as well as monoclonal transgenic γδTCRs can drive the production of both DP and γδTCR⁺ thymocytes in basal states, without any experimental manipulations of TCR signaling. At the population level, both γδTCR⁺ and αβ lineage DP thymocytes are generated in mice lacking Tcrb (and hence the preTCR). The latter subset is absolutely dependent on the γδTCR for DN-to-DP progression [22, 23], but once at the DP stage Tcrg gene transcription is silenced and further maturation ceases. Expression of a transgenic TCRγ chain in Tcrb^−/− mice greatly enhances DP cell generation, but has no significant effect on γδTCR⁺ thymocyte production [23]. Based on similar total numbers and proliferative rates of γδTCR⁺ and DP thymocytes in Tcrb^−/− mice [23], about half of γδTCRs would appear to be weak signalers that specify αβ T cell fate. Hence, at the population level, a large number of γδTCRs appear to lack the capacity to generate lineage discriminatory signals. Conversely, it could be argued that the preTCR does not exclusively trigger precursor cells to differentiate into αβ DP cells, as suggested by the finding that ~15% of γδ TCR⁺ cells harbor a functional TCRβ chain [24].

One way out of this quandary for the instructive model is to introduce an element of stochasticity into the TCR/ligand-based interaction upon which instruction of a progenitor cell occurs. Thus, a given γδTCR can experience several possible outcomes. One is failure (within a prescribed developmental window) to transmit a cognate ligand-based signal, leading to dependence on weak tonic signals (from dimerization [16] or mere delivery of the TCR to the cell surface) for survival and differentiation to DP cells. Alternatively, the γδTCR may transmit a strong signal in response to non-cognate intrathymic ligands. For instance, some T22-specific γδT cells, such as the KN6 TCR, may mature into γδT cells in response to strong signals from interaction with non-cognate MHC-like ligands (some of which must be β2m independent, as T10/22-reactive γδ T cells are produced in B2m^−/− mice). Finally, the γδTCR may respond to a different set of non-cognate ligands with extremely high affinity leading to cell death (for example, the G8 γδTCR recognizes T22 with ~100 times higher affinity than KN6 and is subject to negative selection [15, 25]). Experimentally decreasing signal strength will increase the generation of DP cells, perhaps at the expense of γδ T cells, with the caveat that those encounters that normally negatively select γδ T cells can be tuned lower to positively select γδ T cells. Hence, the quantitative instructive model can be upheld, ad hoc, by invoking stochastic encounters with a variety of potential ligands that elicit a spectrum of signal strengths, even from a single γδTCR.

However, a similar speculative logic can turn these observations to support the stochastic model. αβ T cell lineage-biased precursors expressing the γδTCR can become DP cells with the similar conditions as above, most important being that DP cells will be negatively selected for those experiencing the stochastic TCR signaling at the higher end of spectrum. Turning down the capacity to signal will increase DP cell numbers by increasing the negative selection signal threshold and thereby decreasing the frequency of interactions that result in negative selection. It is also possible that γδTCRs expressed in αβ lineage committed cells prior to the DP stage (particularly prevalent in γδTCR Tg mice) will lead to cell death. This type of transgenic artifact has been demonstrated in the H-Y αβTCR Tg system, where the decrease in DP thymocyte numbers has been convincingly shown to arise from the premature expression of the αβTCR transgene in DN cells [26]. By attenuating the ability of γδTCRs to signal, the death of DPs may be prevented, thereby increasing the size of the DP compartment. At the same time, γδTCRs expressed on γδ lineage biased cells with impaired signaling capacity (below the tonic level) may result in aborted differentiation, resulting in a reciprocal loss in γδ cell numbers.

4. TCR signal strength and γδ T cell subset generation

γδ T cells can be segregated based on Tcrg gene usage, and the subsets thus defined exhibit distinct activation profiles and effector functions. These γδ T cell subsets have different TCR signaling requirements for development and maintenance [16, 27, 28]. Whether these differences are relevant when γδTCRs are first expressed on DN precursor cells is unknown. However, the very distinct gene expression profiles of γδ T cell subsets (for example, previously identified γδ lineage biased genes such as Rgs2, Crem and Bcl2a1 [29] are primarily expressed in Vγ1.1⁺Vδ6.3⁺ NKT cells, unpublished and Immgen.org), the attendant acquisition of unique effector functions from the earliest stages of γδTCR⁺ thymocyte maturation, and the unknown nature of ligands that participate in speculative γδ T cell subset-specific selection processes minimally warrant caution against assuming that a relatively narrow spectrum of signaling (i.e., strong) is imparted by different γδTCRs.

PLZF (Zbtb16) is expressed exclusively in innate-like αβNKT, MAIT (innate-like cells that are found in the gut lamina propria), and γδNKT cells [30–33]. γδNKT cells express a specific TCR consisting of Vγ1.1 paired with Vδ6.3 [34]. Like αβNKT cells, they are found in the thymus and peripheral lymphoid organs, but they are most abundant in the liver [35, 36]. Based on the selective expansion of γδNKT cells in mice deficient in Itk [37, 38], Id3 [33], and with a mutation in Slp76 [33] and Lat [39] (and likely in Cd3z^−/− mice [40]), it appears that decreasing the TCR signal strength expands the γδNKT cell lineage. While one study showed a reduction in the number of other Vγ expressing cells in Id3^−/− mice [41], we and others [33, 42, 43] have not noted any differences in the number of γδ T cells other than Vγ1.1⁺Vδ6.3⁺ T cells in the mutant mice, indicating that there was no skewing away from the γδ lineage as a whole. Interestingly, while approximately 70% of wt γδNKT cells express PLZF, nearly all γδNKTs from Id3^−/− [33], Itk^−/− (unpublished) and Slp-76 [33] mutant mice express PLZF, even though all these models result in reduced TCR signaling. In a parallel fashion, Itk and Slp76 mutations led to an increased production of innate-like CD8⁺ or CD4⁺ αβ T cells [44, 45]. Therefore, the prediction that decreasing the TCR signal strength globally will generate more αβ T cells at the expense of γδ T cells at the population level is clearly not supported by the data. Rather, decreasing TCR signal strength appears to generate more innate-like lymphocytes expressing the αβ or γδTCR.

Another interpretation of these data in the context of the TCR signal strength model is that relatively strong TCR signals generate γδNKT cells, and that a negative selection threshold exists for γδNKT cells such that too much signal normally deletes them. Consistent with this possibility, Id3^−/− mice exhibit defects in negative selection in TCR Tg mice [46]. However, while γδTCR-triggering in vitro (strong signal, using anti-TCRδ chain Ab in the OP9-Dl1 culture system) exclusively generates PLZF⁺ T cells [32], so far it has not been possible to delete these cells using stronger γδTCR triggering, if such a condition exists. Furthermore, consistent with a pre-existing heterogeneity in signal interpretation, only about half of γδ T cells in Vg1.1/Vd6.4 (Vδ6.3 allele of DBA/J mice, [34]) TCR transgenic mice are PLZF⁺ [32], despite the fact that the cells all have an identical TCR and should receive the same signal strength. If signal strength is not determining PLZF expression, it seems likely that other molecular differences among γδNKT cells may determine the proportion of cells that express PLZF. Alternatively, the ad hoc hypothesis that a clonal γδTCR can transmit varied signals can be invoked to explain this observation.

In addition to these contradictions with the signal strength model, there are several other confounding issues for the role of TCR signal strength in γδNKT cell generation. First, γδNKT cell development is dependent on SAP [32, 33], which is required for the development of αβNKT [47–49] cells and other innate-like αβ T cells [50] via SLAM protein interactions, but not for “conventional” αβ or γδ T cells [51]. Hence, there are certain to be qualitative differences in signaling, and intrathymic cellular interactions, regulating the generation of distinct T cell subsets. Second, genes that specify lineage function such as Plzf, are expressed early in DN1/2 cells (Immgen.org), potentially programming lineage function prior to TCR expression. Third, although PLZF is only thought to be functional in NKT cells, Plzf^−/− mice have reduced numbers of total γδ thymocytes [33], not just the γδNKT cells, suggesting that PLZF expression in early precursor cells may be developmentally relevant or that PLZF⁺ cells impact other T cells in trans. Fourth, the nature of intrathymic γδTCR ligands positively or negatively selecting γδ T cells or γδNKT cells is, for the most part, unknown, making it difficult to design experimental probes to test the model.

In summary, it is clear that decreasing TCR signal strength has a profound impact on the generation of one subset of γδ T cells that belong to the NKT cell lineage and on innate-like αβ T cells, while having no apparent effects on the generation of other “conventional” γδ and αβ lineage cells at the population level. Hence, the γδTCR signal spectrum is not digital and has the potential capacity to shape the γδTCR repertoire. It is very plausible that the γδNKT cell subset expansion in the TCR signaling mutant mice arises from qualitative differences in signaling or clonal variations in lineage potential that interpret the same TCR signals distinctly. Identification of ligands that impact different γδ T cell subsets will be necessary to definitively resolve this issue.

5. Evidence that γδ/αβ lineage choice is stochastic

5.1 DN2 precursors are molecularly heterogeneous

The body of evidence that TCR lineage commitment is stochastic is also not fully formed. Molecular heterogeneity at the TCR⁻ DN2 stage is a necessary condition of the stochastic model that has been proven to exist. Cytokine receptors (IL-7Rα [7], Activin RII, unpublished), preTCRα [52], transcription factors (SOX13 [53] and SOX5, unpublished), and the components of the WNT signaling pathway (AXIN2, [54] and unpublished) are a few of the factors differentially expressed in single DN2 cells. There are likely many more, given that several genes that are uniquely expressed in γδ cells are potentially only expressed in a subset of DN2 cells. The most direct proof of the stochastic model is the purification of putative lineage committed precursors that can give rise to only γδ or αβ T cells. Given the paucity of cell surface markers that can be used to segregate DN2 cells the first step is to generate a panel of mice containing DN2 cells that are differentially marked, in most cases by placing a reporter gene (Gfp or LacZ, for example) under the control of promoters that are variably active among single DN2 cells. Studies using these mice are beginning to yield evidence pointing toward a pre-existing T cell lineage bias prior to TCR expression.

5.2 Developmental potential of single DN2 cells

Ciofani et. al provided the first evidence at the single cell level that the major point of lineage commitment occurs at the DN2/DN3 stage [55]. They sorted single DN1, DN2, and DN3 cells and tested their developmental potential using the OP9-DL1 culture system. The results showed that DN1 cells were not committed to either the αβ or γδ lineage, as almost all single cells were capable of giving rise to both αβ and γδ T cells. In contrast, over half of DN2 cells were already committed to one T cell lineage: from the fetus ~60% of DN2 cells were capable of generating only αβ T cells, while ~3% generated only γδ T cells; in comparison ~25% and ~20% of adult DN2 cells were capable of generating only αβ and γδ T cells, respectively [55]. While the results from the adult DN2 cells support the existence of lineage committed cells prior to TCR gene rearrangement, they also point to a difference in developmental potential between fetal and adult DN2 cells. It has been noted previously that adult and fetal T cell development requirements are distinct [56, 57]. In addition, it is possible that the OP9-DL1 culture system, which provides high levels of Notch signaling and IL-7, distinctly influences fetal versus adult T cell development from DN2 cells.

While the success rate of the gene rearrangement process (one productive TCR gene rearrangement required for the preTCR versus two for γδ T cells) alone has been proposed to account for the γδ and αβ T cell precursor frequencies in the culture, this parameter cannot account for the adult DN2 developmental potential, as the single cells generated only αβ or γδ T cells with nearly equal frequencies [55]. It also can not explain the surprising observation that fetal DN2 cells are exceedingly poor at γδTCR-mediated lineage determination in the OP9 system, given that in the fetus there is a temporal order to TCR gene rearrangements with the Tcrg and d loci rearrangement occurring prior to the b loci (also evident in adults, [1, 2]), which significantly alters the rate of generation of productive γδTCR versus preTCR. In vivo, if the competitive advantage is a key element of the TCR-mediated lineage instruction, then removing the advantage by introducing a functional TCRγ transgene (in otherwise wt mice) expressed early in DN2/DN3 cells should decrease the size of the αβ T cell pool. This, too, was not observed, as the numbers of DP thymocytes were unaltered while γδ cell numbers increased by 2–3 fold in Tcrg Tg mice [23]. Results from other manipulations with similar intent, including Tcrb Tg mice [58, 59], were suggestive of two pools of cells that were generated in parallel from two sources of precursor cells rather than from a common source with two mutually exclusive, competing fates. Hence, the TCR gene rearrangement process per se as the primary driver of precursor cell fate cannot be easily sustained.

Despite the uncertainty pertaining to how single precursor cells can give rise to only γδ or αβ cells, the single precursor cell assay in vitro indicated that lineage commitment in adults occurs partly at the DN2 stage, prior to TCR expression, and is completed by the DN3 stage at which point most cells are αβ lineage restricted.

5.3 Developmental potential of DN2 cells segregated by IL-7R

About half of DN2 cells express relatively high levels of IL-7R on their cell surface. Intrathymic injection of IL-7R^hi DN2 cells favored γδ T cell development, while injection of IL-7R^lo DN2 cells favored αβ T cell lineage development, showing that lineage biased precursors do exist prior to TCR expression [7]. Given that IL-7R signaling has a profound and selective influence on Tcrg locus rearrangement [60] and expression [61, 62], it was possible that a difference in TCR rearrangement efficiency could be responsible for the biased development of precursors distinguished by IL-7R expression. However, even when functional Tcrg gene transcription was enforced equally in both DN2 subsets, the lineage bias was observed [7]. While rearranged TcrVd5 transcripts were detectable in the IL-7R^hi DN2 cell subsets, this is a rare event (<1% based on results in [1]). These findings suggested that DN2 cells are not a homogeneous population and prompted studies to identify other genes of developmental-relevance that are differentially expressed among DN2 cells.

5.4 Developmental potential of Sox13⁺ DN2 cells

Sox13 was the first γδ lineage-restricted gene identified and is also a candidate T cell lineage fate determinant as it is expressed in half of DN2 cells as determined by single cell (sc) RT-PCR [53]. As γδ T cells mature in the thymus and periphery, Sox13 expression decreases, a pattern that is shared by many γδ lineage-restricted transcription factors (Immgen.org). While all CD24⁺ γδ immature thymocytes express Sox13, as assessed by scRT-PCR, Sox13 expression is quantitatively heterogeneous in γδ T cell subsets defined by the V gene usage, with Vγ2⁺ cell subsets having the highest relative expression in the thymus and periphery (unpublished and Immgen.org). NK1.1⁺PLZF⁺ γδNKT cells do not express Sox13, a finding consistent with a previous report [17]. This complex pattern of Sox13 expression has caused some to suggest that it is not a pan γδ T cell marker and not essential for all γδ T cells. As alluded to in section 4, it is not at all clear that peripheral γδ T cells can be defined as one lineage since gene expression profiles of different γδ T cell subsets are as different from each other as they are from the expression profiles of αβ lineage CD8⁺ thymocytes (Immgen.org and unpublished). Nevertheless, Sox13 expression in all CD24⁺ γδ thymocytes still positions it as the best candidate molecule to influence T cell lineage commitment prior to TCR expression. SOX13 may function like NOTCH signaling, which is critical for DN precursor differentiation, but not essential for all peripheral αβ T cell subsets. Once the lineage is determined, Sox13 may have additional roles in maintaining specific γδ T cell subsets. One possible model is that while Sox13 expression is necessary for γδ T cell development, the relative amount of Sox13 and/or the presence of alternative factors regulates the generation of distinct γδ T cell sublineages defined by V gene usage.

In mice, Sox13 promotes γδ T cell development at the expense of the αβ T cell lineage. Consistent with the noted bias toward γδ T cell generation, IL-7R^hi DN2 cells express more Sox13 than IL-7R^lo cells [53]. SOX13 belongs to the SRY-related High Mobility Group (HMG) transcription factor family that contains 25 members. Among the HMG proteins are TCF1 (encoded by Tcf7) and LEF1, which are absolutely essential for αβ T cell development [63, 64], and TOX, which is required for CD4⁺ αβ T cell differentiation from DP thymocytes [65]. SOX13 binds to TCF1 and LEF1 and alters their function in regulating gene transcription [53]. Sox5, a close relative of Sox13, is also expressed in γδ T cells. However, based on analyses of Sox5-reporter mice (from V. Lefebvre, Lerner Research Institute, Cleveland) only ~15% of γδ T cells express Sox5. In DN2 cells, only 5–10% express Sox5 (unpublished), again indicating molecular heterogeneity in the precursor cell subset. Currently, Sox13-reporter mice are being generated and a definitive test of SOX13⁺ DN2 cells as the committed precursors of γδ cells should be forthcoming.

5.5 Developmental potential of DN2 cells segregated by WNT signaling

The most compelling finding so far is that the expression of WNT signaling target genes is also heterogeneous in DN2 cells. The canonical WNT signaling pathway involves β-catenin-TCF1/LEF1 complex-mediated gene regulation. As TCF1 is clearly necessary for normal T cell development [63, 64], the relatively normal hematopoietic lineage development in mice reconstituted with β-Catenin/γ-Catenin double deficient progenitors was a surprise [54, 66]. The implications of this observation for the assumed importance of canonical WNT signaling for T cells are unclear since a redundant factor(s) may exist for Catenins in lymphocytes. Consistent with this possibility, mice deficient in other dedicated downstream signaling molecules of canonical WNT signaling have severe defects in T cell development (unpublished), indicating that WNTs are in fact essential. AXIN2 is a component of the β-catenin destruction complex that is induced by canonical WNT signaling [54, 67]. DN2 cells can be divided into AXIN2^neg (25–30%) and AXIN2^hi (30–40%) populations. Initial results from the examination of the T cell lineage potential pre-existing in DN2 cells distinguished by AXIN2 expression suggest that DN2 cells that can exclusively give rise to γδ T cells can be identified and purified. These results support the contention that the canonical WNT signaling regulated by SOX13-TCF1 complexes is a high value candidate for influencing T cell lineage commitment prior to TCR expression

Other γδ T cell lineage-specific gene reporter mice are in production and simultaneous use of multiple markers to track the precursor-product relationship should permit a thorough testing of the stochastic model of T cell lineage commitment. While these data support the idea that heterogeneous expression of transcription factors and other genes at the DN2 stage significantly impacts the outcome of γδ/αβ lineage commitment, the issue of T cell lineage commitment can only be definitively resolved by a rigorous test of molecularly distinct T cell precursor populations in vivo. This ultimate proof has not yet been generated, but with the advent of more versatile flow cytometry methods using multiple reporter gene knock-in mice combined with in vitro single cell clonal assays, it is becoming possible to dissect DN2 cells into multiple distinct populations based on the expression of molecular effectors of key pathways such as WNT signaling and track their cell lineage fate.

6. Conclusion

While favorable momentum for the TCR signal strength model of T cell lineage commitment clearly exists, many questions remain, and definitive tests for the existence of T cell sublineage-committed cells have yet to be completed. While it was appreciated that γδ T cells segregated based on TCR Vγ/δ gene segment usage are equipped with distinct functions that are programmed during intrathymic differentiation, a recent comprehensive interrogation of γδ sublineage gene expression patterns has revealed extensive differences in gene expression among the different γδ subsets, with at least one subset (PLZF⁺Vγ1.1⁺Vδ6.3⁺) sharing characteristics with the αβNKT cell lineage. This finding questions the very nature of the γδ T cell lineage and challenges the prevailing assumption that distinct TCR signal strengths or master cell fate regulators generate and define all γδ T cells.

Acknowledgements

The authors have been supported by a grant from NIH/NCI (CA100382). We thank members of the lab for discussions and M. Bix for critical comments on the manuscript.

Footnotes

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. Capone M, Hockett RD, Zlotnik A. Kinetics of T cell receptor β, γ, and δ rearrangements during adult thymic development: T cell receptor rearrangemnts are present in CD44+CD25+ Pro-T thymocytes. Proc Natl Acad Sci USA. 1998;95:12522–12527. [PubMed]

2. Livák F, Tourigny M, Schatz DG, Petrie HT. Characterization of TCR gene rearrangements during adult murine T cell development. J of Immunol. 1999;162:2575–2580. [PubMed]

3. Hayes SM, Li L, Love PE. TCR signal strength influences alphabeta/gammadelta lineage fate. Immunity. 2005;22:583–593. [PubMed]

4. Haks MC, Lefebvre JM, Lauritsen JP, Carleton M, Rhodes M, Miyazaki T, et al. Attenuation of gammadeltaTCR signaling efficiently diverts thymocytes to the alphabeta lineage. Immunity. 2005;22:595–606. [PubMed]

5. Kang J, Raulet DH. Events that regulate differentiation of αβ TCR⁺ and γδ TCR⁺ T cells from a common precursor. Sem Immunol. 1997;9:171–179. [PubMed]

6. Livak F, Wilson A, MacDonald HR, Schatz DG. αβ Lineage-committed thymocytes can be rescued by the γδ T cell receptor (TCR) in the absence of TCR β chain. European journal of immunology. 1997;27:2948–2958. [PubMed]

7. Kang J, Volkmann A, Raulet DH. Evidence that gammadelta versus alphabeta T cell fate determination is initiated independently of T cell receptor signaling. J Exp Med. 2001;193:689–698. [PMC free article] [PubMed]

8. Ishida I, Verbeek S, Bonneville M, Itohara S, Berns A, Tonegawa S. T-cell receptor γδ and γ transgenic mice suggest a role of a γ gene silencer in the generation of αβ T cells. Proc Natl Acad Sci USA. 1990;87:3067–3071. [PubMed]

9. Turka LA, Schatz DG, Oettinger MA, Chun JJ, Gorka C, Lee K, et al. Thymocyte expression of RAG-1 and RAG-2: termination by T cell receptor cross-linking. Science. 1991;253:778–781. [PubMed]

10. Wilson A, Held W, MacDonald HR. Two waves of recombinase gene expression in developing thymocytes. J Exp Med. 1994;179:1355–1360. [PMC free article] [PubMed]

11. Aifantis I, Pivniouk VI, Gartner F, Feinberg J, Swat W, Alt FW, et al. Allelic exclusion of the T cell receptor beta locus requires the SH2 domain-containing leukocyte protein (SLP)-76 adaptor protein. J Exp Med. 1999;190:1093–1102. [PMC free article] [PubMed]

12. Hayes SM, Love PE. Distinct structure and signaling potential of the gamma delta TCR complex. Immunity. 2002;16:827–838. [PubMed]

13. Brogdon JL, Leitenberg D, Bottomly K. The potency of TCR signaling differentially regulates NFATc/p activity and early IL-4 transcription in naive CD4+ T cells. J Immunol. 2002;168:3825–3832. [PubMed]

14. Singer A, Adoro S, Park JH. Lineage fate and intense debate: myths, models and mechanisms of CD4− versus CD8-lineage choice. Nature Reviews Immunol. 2008;8:788–801. [PMC free article] [PubMed]

15. Schweighoffer E, Fowlkes BJ. Positive selection is not requried for thymic maturation of transgenic γδ T cells. J Exp Med. 1996;183:2033–2041. [PMC free article] [PubMed]

16. Jensen KD, Su X, Shin S, Li L, Youssef S, Yamasaki S, et al. Thymic selection determines gammadelta T cell effector fate: antigen-naive cells make interleukin-17 and antigen-experienced cells make interferon gamma. Immunity. 2008;29:90–100. [PMC free article] [PubMed]

17. Kreslavsky T, Garbe AI, Krueger A, von Boehmer H. T cell receptor-instructed alphabeta versus gammadelta lineage commitment revealed by single-cell analysis. J Exp Med. 2008;205:1173–1186. [PMC free article] [PubMed]

18. Bruno L, Fehling HJ, von Boehmer H. The ab T cell receptor can replace the gd receptor in the development of gd lineage cells. Immunity. 1996;5:343–352. [PubMed]

19. Terrence K, Pavlovich CP, Matechak EO, Fowlkes BJ. Premature Expression of T Cell Receptor (TCR)alphabeta Suppresses TCRgammadelta Gene Rearrangement but Permits Development of gammadelta Lineage T Cells. J Exp Med. 2000;192:537–548. [PMC free article] [PubMed]

20. Porritt HE, Gordon K, Petrie HT. Kinetics of steady-state differentiation and mapping of intrathymic-signaling environments by stem cell transplantation in nonirradiated mice. J Exp Med. 2003;198:957–962. [PMC free article] [PubMed]

21. Punt J, Osborne B, Takahama Y, Sharrow S, Singer A. Negative selection of CD4+CD8+ thymocytes by T cell receptor-induced apoptosis requires a costimulatory signal that can be provided by CD28. J Exp Med. 1994;179:709–713. [PMC free article] [PubMed]

22. Passoni L, Hoffman ES, Kim S, Crompton T, Pao W, Dong MQ, et al. Intrathymic δ selection events in γδ cell development. Immunity. 1997;7:83–95. [PubMed]

23. Kang J, Coles M, Cado D, Raulet DH. The developmental fate of T cells is critically influenced by TCRgammadelta expression. Immunity. 1998;8:427–438. [PubMed]

24. Wilson A, MacDonald HR. A limited role for beta-selection during gamma delta T cell development. J of Immunol. 1998;161:5851–5854. [PubMed]

25. Adams EJ, Strop P, Shin S, Chien YH, Garcia KC. An autonomous CDR3delta is sufficient for recognition of the nonclassical MHC class I molecules T10 and T22 by gammadelta T cells. Nature immunology. 2008;9:777–784. [PMC free article] [PubMed]

26. Baldwin TA, Sandau MM, Jameson SC, Hogquist KA. The timing of TCR alpha expression critically influences T cell development and selection. J Exp Med. 2005;202:111–121. [PMC free article] [PubMed]

27. O'Brien RL, Roark CL, Jin N, Aydintug MK, French JD, Chain JL, et al. gammadelta T-cell receptors: functional correlations. Immunological reviews. 2007;215:77–88. [PubMed]

28. Ribot JC, deBarros A, Pang DJ, Neves JF, Peperzak V, Roberts SJ, et al. CD27 is a thymic determinant of the balance between interferon-gamma- and interleukin 17-producing gammadelta T cell subsets. Nature immunology. 2009;10:427–436. [PubMed]

29. Pennington DJ, Silva-Santos B, Shires J, Theodoridis E, Pollitt C, Wise EL, et al. The interrelatedness and interdependence of mouse T cell receptor gammadelta+ and alphabeta+ cells. Nature immunology. 2003;4:991–998. [PubMed]

30. Savage AK, Constantinides MG, Han J, Picard D, Martin E, Li B, et al. The transcription factor PLZF directs the effector program of the NKT cell lineage. Immunity. 2008;29:391–403. [PMC free article] [PubMed]

31. Kovalovsky D, Uche OU, Eladad S, Hobbs RM, Yi W, Alonzo E, et al. The BTB-zinc finger transcriptional regulator PLZF controls the development of invariant natural killer T cell effector functions. Nature immunology. 2008;9:1055–1064. [PMC free article] [PubMed]

32. Kreslavsky T, Savage AK, Hobbs R, Gounari F, Bronson R, Pereira P, et al. TCR-inducible PLZF transcription factor required for innate phenotype of a subset of gammadelta T cells with restricted TCR diversity. Proceedings of the National Academy of Sciences of the United States of America. 2009;106:12453–12458. [PubMed]

33. Alonzo ES, Gottschalk RA, Das J, Egawa T, Hobbs RM, Pandolfi PP, et al. Development of promyelocytic zinc finger and ThPOK-expressing innate gammadelta T cells is controlled by strength of TCR signaling and Id3. J Immunol. 184:1268–1279. [PubMed]

34. Azuara V, Levraud JP, Lembezat MP, Pereira P. A novel subset of adult gamma delta thymocytes that secretes a distinct pattern of cytokines and expresses a very restricted T cell receptor repertoire. Eur J Immunol. 1997;27:544–553. [PubMed]

35. Pereira P, Hermitte V, Lembezat MP, Boucontet L, Azuara V, Grigoriadou K. Developmentally regulated and lineage-specific rearrangement of T cell receptor Valpha/delta gene segments. Eur J Immunol. 2000;30:1988–1997. [PubMed]

36. Lees RK, Ferrero I, MacDonald HR. Tissue-specific segregation of TCRgamma delta+ NKT cells according to phenotype TCR repertoire and activation status: parallels with TCR alphabeta+NKT cells. Eur J Immunol. 2001;31:2901–2909. [PubMed]

37. Felices M, Yin CC, Kosaka Y, Kang J, Berg LJ. Tec kinase Itk in gammadeltaT cells is pivotal for controlling IgE production in vivo. Proceedings of the National Academy of Sciences of the United States of America. 2009;106:8308–8313. [PubMed]

38. Qi Q, Xia M, Hu J, Hicks E, Iyer A, Xiong N, et al. Enhanced development of CD4+ gammadelta T cells in the absence of Itk results in elevated IgE production. Blood. 2009;114:564–571. [PubMed]

39. Nunez-Cruz S, Aguado E, Richelme S, Chetaille B, Mura AM, Richelme M, et al. LAT regulates gammadelta T cell homeostasis and differentiation. Nature immunology. 2003;4:999–1008. [PubMed]

40. Arase H, Ono S, Arase N, Park SY, Wakizaka K, Watanabe H, et al. Developmental arrest of NK1.1+ T cell antigen receptor (TCR)-alpha/beta+ T cells and expansion of NK1.1+ TCR-gamma/delta+ T cell development in CD3 zeta-deficient mice. J Exp Med. 1995;182:891–895. [PMC free article] [PubMed]

41. Lauritsen JP, Wong GW, Lee SY, Lefebvre JM, Ciofani M, Rhodes M, et al. Marked induction of the helix-loop-helix protein Id3 promotes the gammadelta T cell fate and renders their functional maturation Notch independent. Immunity. 2009;31:565–575. [PMC free article] [PubMed]

42. Verykokakis M, Boos MD, Bendelac A, Adams EJ, Pereira P, Kee BL. Inhibitor of DNA binding 3 limits development of murine slam-associated adaptor protein-dependent "innate" gammadelta T cells. PloS one. 5:e9303. [PMC free article] [PubMed]

43. Ueda-Hayakawa I, Mahlios J, Zhuang Y. Id3 restricts the developmental potential of gamma delta lineage during thymopoiesis. J Immunol. 2009;182:5306–5316. [PMC free article] [PubMed]

44. Prince AL, Yin CC, Enos ME, Felices M, Berg LJ. The Tec kinases Itk and Rlk regulate conventional versus innate T-cell development. Immunological reviews. 2009;228:115–131. [PMC free article] [PubMed]

45. Jordan MS, Smith JE, Burns JC, Austin JE, Nichols KE, Aschenbrenner AC, et al. Complementation in trans of altered thymocyte development in mice expressing mutant forms of the adaptor molecule SLP76. Immunity. 2008;28:359–369. [PMC free article] [PubMed]

46. Rivera RR, Johns CP, Quan J, Johnson RS, Murre C. Thymocyte selection is regulated by the helix-loop-helix inhibitor protein, Id3. Immunity. 2000;12:17–26. [PubMed]

47. Pasquier B, Yin L, Fondaneche MC, Relouzat F, Bloch-Queyrat C, Lambert N, et al. Defective NKT cell development in mice and humans lacking the adapter SAP, the X-linked lymphoproliferative syndrome gene product. J Exp Med. 2005;201:695–701. [PMC free article] [PubMed]

48. Chung B, Aoukaty A, Dutz J, Terhorst C, Tan R. Signaling lymphocytic activation molecule-associated protein controls NKT cell functions. J Immunol. 2005;174:3153–3157. [PubMed]

49. Nichols KE, Hom J, Gong SY, Ganguly A, Ma CS, Cannons JL, et al. Regulation of NKT cell development by SAP, the protein defective in XLP. Nature medicine. 2005;11:340–345. [PubMed]

50. Horai R, Mueller KL, Handon RA, Cannons JL, Anderson SM, Kirby MR, et al. Requirements for selection of conventional and innate T lymphocyte lineages. Immunity. 2007;27:775–785. [PMC free article] [PubMed]

51. Schwartzberg PL, Mueller KL, Qi H, Cannons JL. SLAM receptors and SAP influence lymphocyte interactions, development and function. Nature reviews. 2009;9:39–46. [PubMed]

52. Gounari F, Aifantis I, Martin C, Fehling HJ, Hoeflinger S, Leder P, et al. Tracing lymphopoiesis with the aid of a pTalpha-controlled reporter gene. Nature immunology. 2002;3:489–496. [PubMed]

53. Melichar HJ, Narayan K, Der SD, Hiraoka Y, Gardiol N, Jeannet G, et al. Regulation of gammadelta versus alphabeta T lymphocyte differentiation by the transcription factor SOX13. Science. 2007;315:230–233. [PubMed]

54. Jeannet G, Scheller M, Scarpellino L, Duboux S, Gardiol N, Back J, et al. Long-term, multilineage hematopoiesis occurs in the combined absence of beta-catenin and gamma-catenin. Blood. 2008;111:142–149. [PubMed]

55. Ciofani M, Knowles GC, Wiest DL, von Boehmer H, Zuniga-Pflucker JC. Stage-specific and differential notch dependency at the alphabeta and gammadelta T lineage bifurcation. Immunity. 2006;25:105–116. [PubMed]

56. Kang J, DiBenedetto B, Narayan K, Zhao H, Der SD, Chambers CA. STAT5 is required for thymopoiesis in a development stage-specific manner. J Immunol. 2004;173:2307–2314. [PubMed]

57. Wang JH, Nichogiannopoulou A, Wu L, Sun L, Sharpe AH, Bigby M, et al. Selective defects in the development of the fetal and adult lymphoid system in mice with an Ikaros null mutation. Immunity. 1996;5:537–549. [PubMed]

58. Gerber D, Boucontet L, Pereira P. Early expression of a functional TCRbeta chain inhibits TCRgamma gene rearrangements without altering the frequency of TCRgammadelta lineage cells. J Immunol. 2004;173:2516–2523. [PubMed]

59. Krotkova A, Smith E, Nerz G, Falk I, Eichmann K. Delayed and restricted expression limits putative instructional opportunities of Vgamma1.1/Vgamma2 gammadelta TCR in alphabeta/gammadelta lineage choice in the thymus. J Immunol. 2004;173:25–32. [PubMed]

60. Maki K, Sunaga S, Ikuta K. The V-J recombination of T cell receptor-gamma genes is blocked in interleukin-7 receptor-deficient mice. J Exp Med. 1996;184:2423–2427. [PMC free article] [PubMed]

61. Kang J, Coles M, Raulet DH. Defective development of γδ T cells in interleukin 7 receptor-deficient mice is due to impaired expression of T cell receptor γ genes. J Exp Med. 1999;190:973–982. [PMC free article] [PubMed]

62. Perumal NB, Kenniston TW, Jr, Tweardy DJ, Dyer KF, Hoffman R, Peschon J, et al. TCR-gamma genes are rearranged but not transcribed in IL-7R alpha-deficient mice. J Immunol. 1997;158:5744–5750. [PubMed]

63. Verbeek S, Izon D, Hofhuis F, Robanus-Maandag E, te Riele H, van de Wetering M, et al. An HMG-box-containing T-cell factor required for thymocyte differentiation. Nature. 1995;374:70–74. [PubMed]

64. Okamura RM, Sigvardsson M, Galceran J, Verbeek S, Clevers H, Grosschedl R. Redundant regulation of T cell differentiation and TCRalpha gene expression by the transcription factors LEF-1 and TCF-1. Immunity. 1998;8:11–20. [PubMed]

65. Aliahmad P, Kaye J. Development of all CD4 T lineages requires nuclear factor TOX. J Exp Med. 2008;205:245–256. [PMC free article] [PubMed]

66. Koch U, Wilson A, Cobas M, Kemler R, Macdonald HR, Radtke F. Simultaneous loss of beta- and gamma-catenin does not perturb hematopoiesis or lymphopoiesis. Blood. 2008;111:160–164. [PubMed]

67. Lustig B, Jerchow B, Sachs M, Weiler S, Pietsch T, Karsten U, et al. Negative feedback loop of Wnt signaling through upregulation of conductin/axin2 in colorectal and liver tumors. Mol Cell Biol. 2002;22:1184–1193. [PMC free article] [PubMed]