Overexpression of RhoH Permits to Bypass the Pre-TCR Checkpoint

RhoH, an atypical small Rho-family GTPase, critically regulates thymocyte differentiation through the coordinated interaction with Lck and Zap70. Therefore, RhoH deficiency causes defective T cell development, leading to a paucity of mature T cells. Since there has been no gain-of-function study on RhoH before, we decided to take a transgenic approach to assess how the overexpression of RhoH affects the development of T cells. Although RhoH transgenic (RhoHtg) mice expressed three times more RhoH protein than wild-type mice, β-selection, positive, and negative selection in the thymus from RhoHtg mice were unaltered. However, transgenic introduction of RhoH into Rag2 deficient mice resulted in the generation of CD4+CD8+ (DP) thymocytes, indicating that overexpression of RhoH could bypass β-selection without TCRβ gene rearrangement. This was confirmed by the in vitro development of DP cells from Rag2-/-RhoHtg DN3 cells on TSt-4/Dll-1 stroma in an Lck dependent manner. Collectively, our results indicate that an excess amount of RhoH is able to initiate pre-TCR signaling in the absence of pre-TCR complexes.


Introduction
T cells develop in the thymus through a complex multistage process. It is well known that two major checkpoints exist during T cell development in the thymus. The first checkpoint is βselection (or pre-TCR checkpoint) at the CD4 -CD8 -CD25 hi CD44 low (DN3) stage, and the other is repertoire selection (positive and negative selection) at the CD4 + CD8 + double positive (DP) stage. Progression through both stages is dependent on pre-TCR or TCR complexes, therefore many TCR signal-related molecules are involved in these checkpoints [1,2].
RhoH belongs to the Rho family small GTPases, which play crucial roles in the development of thymocytes [3]. RhoH is an atypical small G protein that lacks GTPase activity, and is expressed exclusively in hematopoietic lineage cells [4,5]. We and others have demonstrated that RhoH acts as an adaptor protein associating with the Src-family protein tyrosine kinase Lck, c-Src tyrosine kinase Csk, and Syk-family protein tyrosine kinase Zap70 and Syk [6][7][8]. Because RhoH is anchored to the plasma membrane through myristoylation, it is able to recruit Lck, ZAP70, and Syk to the membrane to facilitate membrane proximal signal transduction [6,7,9]. Ligation of TCR by MHC/peptide complexes induces phosphorylation of TCR-z chain ITAMs by Lck, thereby initiating TCR-dependent signal transduction. Zap70 is recruited to these phosphorylated Immunoreceptor tyrosine-based activation motifs (ITAMs), where Zap70 is phosphorylated and converts to the catalytically active form that phosphorylates SLP76 and LAT which then transduce signals to their downstream targets. Since RhoH controls these two proximal kinases important for the initiation of TCR signaling, it is crucial for T cell development and activation. Consequently, TCR-mediated signal transduction in RhoH deficient T cells is almost completely abolished, resulting in defective T cell development through impaired β-selection, positive selection, and negative selection [4,5,10].
β-selection, one of the two major checkpoints during T cell development, ensures successful production of a TCRβ chain. Therefore, mice having defective gene rearrangement, such as Rag1/2 deficiency or scid mutation, could not pass beyond the checkpoint, resulting in a strict developmental arrest at the DN3 stage. However, the following treatments have been shown to allow checkpoint bypass without TCRβ protein synthesis: irradiation-induced DNA damage [11,12], inhibition of apoptosis [13,14], or forced activation of pre-TCR signaling [15][16][17]. Indeed, constitutively active mutations in signaling molecules such as Lck, Ras and Rac could also induce DN-DP transition independent of TCRβ gene rearrangement [16,18,19].
Besides a gene knockout approach, gain-of-function analysis, as typified by transgenic overexpression, is also useful to explore gene function. In this article, we have investigated the function of RhoH by utilizing RhoH transgenic (RhoH tg ) mice under the control of the CD2 promoter. Overexpression of RhoH in thymocytes did not interfere with T cell development significantly, however, we found that excess RhoH resulted in bypass of the β-selection checkpoint, allowing differentiation from DN to DP without TCRβ recombination in vivo. This DN to DP transition was also observed by in vitro culture, and was dependent on Lck. Since RhoH facilitates activation of Lck and Zap70, excess amounts of RhoH protein could initiate Lckdependent activation of pre-TCR signaling in the absence of pre-TCR complexes.

Ethics statements
Animal experiments were approved by the Animal Care and Use Committee of the National Center for Global Health Medicine (NCGM) Research Institute and conducted in accordance with institutional procedures. (#14033) All efforts were made to minimize suffering.

Animals
RhoH tg mice were generated by the microinjection of VA hCD2-HA-RhoH vector into the pronuclei of fertilized eggs using standard procedures. Previous studies have shown VA vector containing human CD2 promoter/T cell specific transcriptional enhancer [20] directs the expression of transgenes in mice to the T cell lineage [21]. The embryos were transferred to the oviducts of pseudopregnant ICR female mice. Established mouse lines were maintained after 10 generations of backcrossing to C57BL/6J (B6), and housed under specific pathogen-free conditions in accordance with institutional guidelines. Mice were sacrificed by cervical dislocation to dissect organs out.

Statistical analysis
Statistical analyses were performed using GraphPad Prism statistical analysis software. Group differences were analyzed by unpaired Student's t-test or two-way ANOVA with multiple comparisons, followed by Tukey's post-test comparisons, for three or more groups. The p values 0.05 were considered significant.

HA-tagged RhoH transgene restored normal thymocyte differentiation in RhoH deficient mice
We established hCD2 promoter-driven RhoH tg mice that expressed an N-terminal HA-tagged RhoH protein. To confirm whether HA-tagged RhoH protein could function normally, RhoH tg mice were bred to RhoH -/mice. In the RhoH -/-RhoH tg mice, protein expression of RhoH in total thymocytes was increased two-fold compared to wild-type mice ( Fig 1A). Like wild-type non-tagged RhoH protein, most of the HA-tagged RhoH protein in RhoH tg thymocytes localized to the plasma membrane (S1A Fig). As has been published before, RhoH deficient mice showed severe inhibition of T cell development (Fig 1B and 1C) with thymic cellularity and the percentage of CD4SP and CD8SP cells being significantly reduced [4,5]. We found that introduction of an HA-tagged RhoH transgene into RhoH -/mice successfully restored cellularity and the differentiation of CD4SP and CD8SP cells (Fig 1B and 1C). The expression of the transgene were able to restore both defective positive and β-selection ( Fig 1B). Expression of CD2 and CD5 on DP thymocytes, which was severely reduced in RhoH -/- [4,5], was restored by the transgene as well (Fig 1D). The HA-tagged RhoH associated with phosphorylated Lck without stimulation, and was phosphorylated and bound to ZAP70 upon TCR-stimulation (S1B Fig), exactly the same as non-tagged endogenous RhoH. Furthermore, impaired TCRinduced phosphorylation of ERK in RhoH -/thymocytes was restored by the HA-tagged RhoH transgene (S1C Fig). Collectively, these results demonstrate that HA-tagged RhoH is functionally equivalent to the endogenous one, indicating that HA-RhoH transgenic mice are a proper model for analyzing the molecular functions of RhoH.

Effects of over-expression of RhoH on T cells in the thymus and peripheral tissues
To determine gain-of-function effects of RhoH on T cell development, we examined HAtagged RhoH tg mice in C57BL/6 background. Although lack of RhoH caused strong inhibition of T cell development [4,5] overexpression of RhoH did not affect overall differentiation of T cells, as the number and proportion of each subset in the thymus were unchanged (Fig 2A and  S2A Fig). GFP expression on DP to CD4SP cells from RhoH tg Nur77-GFP mice, which were able to monitor the strength of TCR signaling in vivo during T cell developments [23], was also the same as the wild-type controls (S3A Fig). We finally observed RhoH transgene made small increase of the phosphorylation of Src family protein which act as the key kinases on T cell development in thymus (Fig 2B), but subpopulations of DN thymocytes were all unaltered in RhoH tg mice (S2C Fig), indicating that β-selection was not affected. Although prominent differences were not observed in the thymus, we noticed a slight but consistent up-regulation of CD2 and CD5 expression in DP cells from RhoH tg mice (Fig 2C). These changes were still observed in MHC -/-(I-Aβ and β 2 m DKO) background (Fig 2D), which is defective in positive and negative selection, indicating that the increase of CD2 and CD5 was independent of any events later than positive selection. Since expression levels of CD2 and CD5 correlate well to the strength of TCR signal [24,25], these results may indicate augmented pre-TCR signaling in RhoH-overexpressing thymocytes. In the periphery, the percentage and numbers of splenic CD4 + and CD8 + T cells from RhoH tg mice were similar to that of wild-type mice (Fig 2E and  S2B Fig). However, frequency and numbers of naïve CD4 and CD8 T cells (CD44 low CD62L hi ) were significantly reduced, and activated/memory T cells were concomitantly increased in RhoH tg mice (Fig 2F and 2G). Collectively, compared to the severe phenotypes of RhoH deficient mice, overexpression of RhoH had little effect on T cells, apart from the increased phosphorylation of Src family kinases in DN3, increased expression of CD2 and CD5 on DP, decreased number of naive T cells, and increased activated/memory T cells in the periphery.
Other than conventional TCRαβ lineage T cells, development of unconventional T cells such as pTregs, NKT cells, and TCRγδ T cells was not changed in the RhoH tg mice (S4A-S4C Fig).
Overexpression of RhoH enables bypass of β-selection in Rag2 -/mice Rag2 deficient mice are unable to complete gene rearrangement of the TCR locus, therefore T cell development stops at the β-selection checkpoint, resulting in complete arrest at the DN3 stage. To our surprise, when we crossed RhoH tg mice with Rag2 -/mice, DP thymocytes emerged in the thymus (Fig 3A). The generation of DP cells associated with increased gene expression of RhoH was in a dose-dependent manner (Fig 3A), because transgene homozygous tg/tg mice generated more DP than transgene heterozygous tg/-mice. Introduction of RhoH transgene did not induce gene rearrangement in the TCRβ locus, because TCRβ protein was Overexpression of RhoH could successfully bypass βselection to generate DPs, however it could not bypass positive selection because differentiation of mature CD4SP and CD8SP cells was not observed in the Rag2 -/-RhoH tg mice (Fig 3A). This was further confirmed by the results from MHC -/-RhoH tg mice, which contained no single positive cells (Fig 3B), proving that simple overexpression of RhoH is not sufficient for bypassing positive selection. Taken together, overexpression of RhoH enabled bypass of β-selection without TCRβ-gene rearrangement, however it was not sufficient to permit bypass of positive selection.

RhoH-induced bypass of β-selection is Lck-dependent
To address whether generation of DPs in Rag -/-RhoH tg mice was thymocyte intrinsic, we next tested using an in vitro differentiation culture system. Sorted DN3 cells from wild-type mice were able to differentiate into DPs in vitro in a week on monolayers of Notch ligand-expressing stromal cells (TSt-4/Dll-1), while DN3 cells from Rag2 -/mice could not (Fig 4A). We found that sorted DN3 cells from Rag2 -/-RhoH tg mice could successfully differentiate into DP cells ( Fig 4A), demonstrating that these DPs were truly differentiated from DN3 cells, and that this transition was thymocyte intrinsic. Furthermore, this in vitro DPs generation was strongly inhibited by the addition of Lck inhibitor (Fig 4B, S6 Fig), indicating that RhoH-transgene induced generation of DPs was dependent on Lck activity.

Discussion
The current report describes, for the first time, the effect of RhoH-overexpression on the development and activation of T cells. Overexpression of RhoH had little impact on T cell development, although lack of the molecule induces severe defects [4,5]. As a matter of fact, the only differences we observed were the slight increase of phospho-Src in DN3, increase of CD2 and CD5 expression on DP thymocytes, and reduction of naive T cells in the periphery with a concomitant increase of activated/memory T cells. Other than these differences, the development of T cells was apparently normal. As has been reported, RhoH acts as an adaptor molecule for Lck and Zap70; kinases important for initiating TCR-dependent signal transduction. Thus, the absence of RhoH causes a severe reduction in TCR signaling. By using RhoH-overexpressing transgenic mice we expected enhanced TCR signal transduction, however, no strong augmentation of TCR-dependent signaling has been observed so far. In addition to the unaltered Nur77-GFP expression during thymic T cell differentiation (S3A Fig), anti-TCR antibody-induced activation of ERK in DP and CD4SP cells from RhoH tg mice was also not increased (S3B Fig), indicating that enhancement of TCR signal transduction was not occurring in DP and CD4-SP from the RhoH tg mice. RhoH is required for proper transduction of TCR-mediated signals; however, the level of RhoH expression in wild-type mice might be sufficient and already saturated regarding the activation of Lck and Zap70. This could explain why the RhoH-overexpressing transgenic mice did not exhibit a strong phenotype.
We observed a slight but consistent increase in CD2 and CD5 expression in DPs from RhoH tg mice (Fig 2C and 2D). Because expression levels of CD2 and CD5 are correlated with the strength of TCR signal [24,25], this phenotype likely represents increased pre-TCR signaling at the DN3 stage. In fact, we have found that Src family protein including Lck were activated under steady state in DN3 cells from RhoH tg mice, although we did not observe increased TCR-signaling in DP cells (Fig 2B and S3B Fig). Since the expression level of endogenous RhoH protein in DN3 cells is much lower than in DP, the amount of RhoH in DN3 may not be saturated, therefore supplementation with exogenous RhoH expression could enhance pre-TCR signaling.
RhoH has been reported to inhibit Rac activity [26], and overexpression of RhoH into murine hematopoietic progenitor cells via retroviral-mediated gene transfer reduced proliferation and chemotaxis, and increased apoptosis [27]. However, we did not observe reduced proliferation of CD4-SP (S4D Fig), nor increased apoptosis of DN3 cells (S3C and S5B Figs) in RhoH-tg mice. Therefore function of RhoH seems to be different in different cell types.
We also observed a slight change in TCRab repertoire in RhoH transgenic mice (S4D Fig). This could be explained by inadequate activation of Lck due to RhoH overexpression. It is known that excessive activation of Lck such as in the constitutively active Lck tg mice suppresses allelic exclusion at the TCRβ gene locus [28,29], resulting in defective T cell development. Therefore, moderate increases in Lck activation may partially disturb allelic exclusion leading to TCR repertoire changes. The other possibility is mediated by IL-7 signaling. IL-7 induces the expression of integrin a 4 b 7 and preferential expansion of naïve CD4T cells, resulting in diversification of TCR repertoire [30,31]. Because RhoH deficiency in human revealed relative specific defect in skin homing b 7 and a 4 b 7 integrin + T cells [32], further studies are needed to elucidate IL-7 dependency on TCR repertoire change in RhoH tg mice.
We show for the first time that overexpression of RhoH led to bypass of the pre-TCR checkpoint in Rag2 -/mice. RhoH-overexpression did not induce gene rearrangement of the TCRβ locus in Rag2 -/mice, as TCRβ protein was not expressed in DP from the Rag2 -/-RhoH tg mice (S5A Fig). Bypassing the pre-TCR checkpoint in β-selection-deficient DN cells can be achieved by three independent mechanisms. The first mechanism is associated with DNA strand breaks. DNA damaging agents such as ionizing irradiation-induced p53-dependent [33] generation of DP in SCID mice [11,34] and Rag2 deficient mice [12,35]. It is thought that DNA damage induces p53-dependent DNA repair pathways that then somehow induce DP generation. The second mechanism is the inhibition of apoptosis in DN3 cells. The inhibition of apoptosis either by overexpression of FADD-dominant negative protein [13], or inactivation of p53, evoked DP progression in SCID [36], Rag2 deficient [37], and CD3γ deficient [38] thymocytes. DN cells that fail to initiate pre-TCR signaling are destined to die via p53-dependent apoptosis, and inhibition of their death resulted in the unexpected survival of dying cells that could proceed to DP stage. It is of note that p53 functions in a positive way in the first mechanism, whereas it acts in negative way in the second mechanism. The third mechanism is substitution or mimicry of pre-TCR signaling. It was shown that administration of anti-CD3-antibody in Rag2 deficient mice induced DPs [15] by emulating pre-TCR signal transduction. After that, transgenic introduction of various signaling related molecules such as active-Lck [16], CD4 [17], active-Rac [19], RasGRP1 [39], active-Ras [18], CD28/B7 [40], Pim-1 [41], Egr1/NF-ATc1 [42,43], Egr3 [44], and Ly49D/DAP12 [45] were reported to bypass the pre-TCR checkpoint in Rag2 or CD3ε deficient mice. We found apoptosis are not altered in RhoH over-expressing DN3 cells (S3C and S5B Figs), suggesting that the second mechanism may be not in the case.
RhoH facilitates TCR/pre-TCR signal transduction by the recruitment of Lck and Zap70, therefore overexpression of RhoH would presumably enhance TCR signaling. It could be that the overexpression of RhoH activates Lck in the absence of TCRβ protein, thus generating DP cells by the third mechanism. Indeed, the in vitro generation of DPs from RhoH tg DN3 cells on stromal cell monolayers was blocked by Lck inhibitor, indicating that this transition was dependent on Lck, which is consistent with normal differentiation of DN to DP cells. Although we could not observe any apparent increase in anti-TCR mAb-stimulated "strong" signal transduction in RhoH tg mature T cells and DP thymocytes (S3B Fig), it could be different in weaker TCR stimulation, or it could depend on developmental stage. Indeed, the slight increase in CD2 and CD5 on DPs, and the bypassing of β-selection, could be an indication of augmented pre-TCR signaling in DN3 stage. Because expression level of endogenous RhoH was lower in DN3 thymocytes, augmentative effects could be seen in DNs. Therefore, we think the increased expression of RhoH in DN3 initiated Lck-dependent pre-TCR signaling even in the absence of TCRβ protein.
Reduced numbers of naive T cells (CD44 low CD62L hi ) of RhoH tg mice implies that RhoH transgene may regulate this subpopulation of peripheral T cells. Because of the paucity of normal mature T cells in RhoH -/mice, it has been difficult to determine the requirement of RhoH in primary peripheral T cells by using systemic knockout mice. To define the effector function of RhoH in peripheral T cells, further analyses using RhoH conditional knockout mice are needed. In summary, we showed that overexpression of RhoH in DN cells enabled bypass of the β-selection checkpoint during thymocyte development.