c-Myc mediates pre-TCR-induced proliferation but not developmental progression

Immature T cells progress through a number of well-defined developmental stages in the thymus. Commitment of bone marrow-derived T-cell progenitors^1-6  to the T-cell lineage requires stimulation of Notch signaling,^2,3,7,8  and results in the up-regulation of CD25 (double-negative [DN] 2 subset). Survival and proliferation of thymocytes at this stage is supported by cytokines such as interleukin-7 (IL-7) and stem cell factor (SCF). The subsequent CD44^-CD25⁺ (DN3) stage is marked by the rearrangement of the T-cell receptor (TCR) β, γ, and δ loci. Productive TCRβ gene rearrangements and synthesis of TCRβ chains lead to the surface assembly of a functional pre-TCR, comprising the TCRβ as well as the invariant pTα chain and CD3 subunits. Signals emanating from the pre-TCR promote survival and proliferation of immature thymocytes as well as their differentiation to the CD4⁺CD8⁺ double-positive (DP) stage, effectively instructing immature thymocytes to the αβ T-cell lineage.⁹ 

Despite recent progress in understanding pre-TCR signaling many questions remain unanswered. It is currently known that the pre-TCR is constitutively localized in plasma membrane glycoprotein-enriched microdomains (GEMs) from where it signals in a cell autonomous manner.^10,11  Proximal events of pre-TCR signaling include the phosphorylation of Lck and Zap70. Assembly of the pre-TCR and activation of this pathway is accompanied by a biphasic calcium mobilization, which appears to be regulated by cytoplasmic IP₃ and plasma membrane store-operated calcium channels (SOCs), resulting in nuclear factor (NF) of activated T cells (NFAT) and NFκB activation.¹²  These findings, however, are not sufficient to explain the multiplicity of events following the onset of pre-TCR signaling. Recent evidence both from loss- and gain-of-function approaches indicates that several other genes and signaling pathways are involved in the pre-TCR checkpoint. These include kinases such c-Fyn,¹³  Csk,¹⁴  and Pim1,¹⁵  and adaptor proteins such as LAT and SLP-76.¹⁶  Several transcription factors were also shown to be essential at this developmental stage, such as Ikaros,¹⁷  E2A,¹⁸  Runx2,¹⁹  and c-Myb.^20,21  Multiple findings indicate the involvement of a number of signaling cascades, including Notch,^22,23  Wnt,²⁴  and Hedgehog.²⁵  The interactions between these pathways, however, the signals that mediate their activation and their orchestration with respect to pre-TCR signaling are currently unknown.

The basic region/helix-loop-helix/leucine zipper (bHLHZip) transcription factor c-Myc has been described to play a role in lymphocyte development. Members of the Myc family (c-Myc, N-Myc, and L-Myc) play an integral role in proliferation, survival, and differentiation of normal and neoplastic cells. Myc binds E-box DNA motifs as a heterodimer with Max, resulting in cell cycle entry²⁶  and transcriptional activation or suppression of genes.^27-30  c-Myc has been implicated in cell proliferation³¹  as well as the control of cell growth.^32-37  Its expression increases rapidly in response to growth factors^38,39  and B-cell receptor (BCR)⁴⁰  or TCR ligation.⁴¹  Immature B and T lymphocytes express both c-Myc and N-Myc, while mature cells express only c-Myc. Assessing the requirement for c-Myc in T-cell development was hampered by the embryonic lethality of c-Myc deficient mice prior to the development of lymphocytes.⁴²  To bypass this problem, Douglas and colleagues⁴³  generated chimeric animals from Myc^-/- embryonic stem (ES) cells and Rag1^-/- blastocysts in which the Rag1^-/- cells cannot contribute to the lymphoid lineages. In their study, Myc^-/- progenitors populated embryonic thymi but had reduced proliferation and failed to develop beyond the late DN stages, leading to the suggestion that c-Myc is essential for development through the pre-TCR checkpoint. c-Myc-deficient cells did not populate adult thymi at all, indicating additional defects at earlier stages of hematopoietic development. More recently, c-Myc has indeed been reported to control the self-renewal of hematopoietic stem cells (HSCs),⁴⁴  and its conditional ablation in the bone marrow favored self-renewal over differentiation of HSCs in the stem cell niche.⁴⁵  The involvement of c-Myc in early hematopoietic development indicates that studying its role at the pre-TCR checkpoint requires conditional animal models that avoid the accumulation of developmental defects resulting from c-Myc deficiency at earlier stages.

Here we report that c-Myc is rapidly up-regulated upon induction of pre-TCR signaling. To characterize the role of c-Myc at the pre-TCR developmental checkpoint we used mice that allow conditional Cre-mediated thymocyte-specific ablation of this protein starting at the DN3 stage.⁴⁶  Our studies indicate that c-Myc is required for the proliferation but not the differentiation or survival signals emanating from the pre-TCR.

To generate the TetObeta transgenic mice, the cDNA encoding the TCRβ chain from the 2B4 hybridoma was inserted as a SalI-ClaI-blunted fragment in the XhoI-EcoRV sites of the TetOSB polylinker.⁴⁷  A 2.9-kb XhoI fragment containing the tet operator, minimal cytomegalovirus (CMV) promoter, rat β-globin intron, TCRβ cDNA, and rat β-globin polyadenylation signal was used for the transgenic mice. Primer pairs for genotyping and deletion polymerase chain reactions (PCRs) were as follows: LckCre transgene 5′-ATCGCTCGACCAGTTTAGT-3′ (forward), 5′-CGATGCAACGAGTGATGA-3′ (reverse). The floxed Myc allele was detected with 5′-GCCCCTGAATTGCTAGGAAGACTG-3′ (forward) and 5′-CCGACCGGGTCCGAGTCCCTATT-3′ (reverse). All mice were kept under specific pathogen-free conditions in the animal facilities of Tufts-New England Medical Center according to protocol no. 49-03 approved by the Institutional Animal Care and Use Committee.

Four-color fluorescence-activated cell-sorter (FACS) staining was performed as described.⁴⁸  Antibodies were from BD PharMingen (San Diego, CA): anti-CD3ϵ-phycoerythrin (PE), -biotin (17A2 and 500A2), anti-B220-CyChrome (RA3.6B2), anti-CD4-fluorescein-5-isothiocyanate (FITC), -CyChrome, -PE, -allophycocyanin (APC), anti-CD8-FITC, -CyChrome, -PE, -APC (53.6.7), anti-TCRβ-PE, -CyChrome (H57), anti-TCRγδ-PE, -biotin (GL3), anti-pan-NK-PE, -biotin (DX5), anti-CD44-FITC, -PE (IM7), anti-CD25-APC (PC61), anti-Gr1-PE, -biotin (RB6.782), anti-CD11b-PE, -biotin (M1/70), and anti-Ter119-PE, -biotin. Biotinylated antibodies were detected with streptavidin-PE, -CyChrome, or -APC. The FITC-Annexin V labeling kit was from BD PharMingen. Intracellular Bcl-2 was analyzed using anti-murine Bcl-2 (monoclonal antibody [mAb] 3F11; BD PharMingen) and purified hamster IgG (antitrinitrophenol; BD PharMingen) as an isotype control. To analyze DN thymocytes, mature cells expressing lineage (lin) markers (CD4, CD8, TCRβ, TCRγδ, CD19, Gr1, Mac1, Ter119, and DX5) were electronically excluded. Intracellular TCRβ staining was performed as published.⁴⁹  FACS analysis was performed on a Cyan flow cytometer (DakoCytomation, Fort Collins, CO) and data were analyzed using FlowJo software (Tree Star, Ashland, OR).

DN thymocytes were enriched by depletion of lin⁺ cells using streptavidin-conjugated magnetic beads (Dynal, Oslo, Norway). Cell sorting was performed on a MoFlo cell sorter (DakoCytomation).

mRNA was extracted from sorted cells using the High-Pure RNA Isolation Kit (Roche, Indianapolis, IN). cDNA from 50 000 cells was prepared with the Superscript-II RT kit (Invitrogen, Carlsbad, CA). Samples were equilibrated with respect to β-actin using SYBR Green quantitative PCR on an OpticonII machine (Bio-Rad, Hercules, CA). Semiquantitative PCR was performed on 1:5 serial dilutions. All PCR amplifications used touchdown conditions reaction volume was 30 μL. Primer pair sequences were (forward, reverse): p53, 5′-CCCGAGTATCTGGAAGACAG-3′, 5′-ATAGGTCGGCGGTTCAT-3′; Bcl-x_L, 5′-AGCAACCGGGAGCTGGTGGTCGAC-3′, 5′-GACTGAAGAGTGAGCCCAGCAGA-3′; TCRβ, 5′-AGCTGAGCTGGTGGGTGAATGG-3′, 5′-CCTCTGGCCACTTGTCCTCCTCTG-3′; pTα, 5′-GGCACCCCCTTTCCGTCTCT-3′,5′-TTTGAAGAGGAGCAGGCGCA-3′; c-Myc, 5′-TCACCAACAGGAACTATGAC-3′, 5′-AAGCTCTGGTTCACCATGTC-3′; N-Myc, 5′-GATGATCTGCAAGAACCCAG-3′, 5′-GGATGACCGGATTAGGAGTG-3′; cyclin D2, 5′-CTTCCAAGCTGAAAGAGACC-3′, 5′-TACCCAACACTACCAGTTCC-3′; cyclin D3, 5′-CGAGCCTCCTACTTCCAGTG-3′, 5′-GGACAGGTAGCGATCCAGGT-3′; cyclin E1, 5′-TCCTGGCTGAATGTCTA-3′, 5′-CTTCTCTATGTCGCACCA-3; and β-actin, 5′-TGGAATCCTGTGGCATCCATG-3′, 5′-TAAAACGCAGCTCAGTAACAG-3′.

Quantitative reverse-transcription PCR (qRT-PCR) was performed in real time using an ABI7300 machine (Applied Biosystems, Foster City, CA). p21^Cip1 and Gadd45α were determined relative to GAPDH expression using TaqMan Gene Expression Assays from Applied Biosystems. c-Myc, Tis21, Id3, and Egr1 were assayed with SYBR Green technology, and expression levels were determined relative to β-actin or as described in “Semiquantitative RT-PCR.” Primer sequences were as follows (forward, reverse): β-actin, 5′-ATGGTGGGAATGGGTCAGAA-3′, 5′-TCTC-CATGTCGTCCCAGTTG-3′; Tis21, 5′-ACGCACTGACCGATCATTACA-3′, 5′-GGCTGGCTGAGTCCAATCTGG-3′; Egr1, 5′-TGAGCACCTGACCA-CAGAGTCC-3′,5′-TGGACGGCACGGCACAGCTCAG-3′; and Id3, 5′-GGCACTGTTTGCTGCTTTAGG-3′,5′-GTAGCAGTGGTTCATGTCGTC-3′. All qRT-PCR reactions were in triplicate in 20 μL volume containing 0.3 μM of each primer. The conditions for all qRT-PCRs were: 50°C for 2 minutes and 95°C for 3 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute.

Pellets of total thymocytes were lysed in RIPA buffer supplemented with Protease Inhibitor Cocktail (Roche) and 1 mM PMSF. Samples were resolved on Bis-Tris gradient gels (Invitrogen) and transferred onto nitrocellulose membranes. Secondary antibodies were conjugated to horse-radish peroxidase (HRP). The signal was detected using the enhanced chemoluminescence Plus (ECL Plus kit; Amersham Biosciences, Arlington Heights, IL).

Sorted cells (10⁶-1.5 × 10⁶) were resuspended in 100 μL PBS/0.1% BSA and 5 μM CFSE and incubated at 37°C for 10 minutes before washing extensively. Viability after labeling exceeded 60%. CFSE-labeled cells were cultured in 10-cm tissue-culture plates containing a confluent monolayer of OP9-DL1 cells. Cocultures were maintained in the presence of 5 ng/mL Flt3L (PeproTech, Rocky Hill, NJ) and 1 ng/mL IL-7 (R&D Systems, Minneapolis, MN) for 4 to 6 days.

We examined the effect of pre-TCR signaling on c-Myc by analyzing a novel mouse strain that allows inducible pre-TCR expression. More specifically, a transgenic strain was generated using a cDNA encoding a TCRβ chain under the control of a minimal CMV promoter augmented with 7 tet operator sequences (tetO). Inducible TCRβ expression was achieved by crossing onto a second strain⁴⁷  that expressed the TetR-VP16 transactivator (tTA) in immature thymocytes under the control of the Lck gene proximal promoter (LTH-1). Expression of the transgenic TCRβ chain in the compound mice is suppressed in the presence and induced in the absence of tetracycline. To avoid simultaneous expression of multiple TCRs the inducible pre-TCR mice were crossed onto the Rag1^-/- background. In the presence of tetracycline the thymic profile of the TetOβ-LTH-Rag1^-/- mice resembled that of Rag1^-/- mice with a thymic cellularity of 2 × 10⁶ to 3 × 10⁶, indicating that the drug effectively suppressed TCRβ transgene expression and prohibited pre-TCR assembly. In the absence of tetracycline the tTA induced the expression of the transgenic TCRβ chain leading to the assembly of a functional pre-TCR and developmental progression to the DP stage. Untreated thymi contained 75% to 80% DP cells and had a thymic cellularity of 50 × 10⁶ to 60 × 10⁶ cells. Thymic lobes isolated from newborn mice treated with tetracycline during gestation were placed in organ cultures for 5 days without tetracycline. This resulted in developmental progression with more than 70% of the thymocytes reaching the DP stage and a more than 10-fold increase in the cellularity of the lobes. All DP cells and the majority of DN4 cells expressed intracellular TCRβ chains (Figure 1A), indicating that they had assembled a functional pre-TCR and undergone beta selection. These newly developing thymocytes present an optimal model system for assessing the downstream effects of pre-TCR signaling.

The effect of pre-TCR signaling on c-Myc was examined after induction of TCRβ expression in TetOβ-LTH-Rag1^-/- thymocytes. To this aim, thymocytes isolated from TetOβ-LTH-Rag1^-/- or LTH-Rag1^-/- animals treated with tetracycline were cocultured with OP9-DL1 stromal cells for 24 hours in tetracycline-free culture medium. Suspension cells were recovered from the cocultures and thymocytes (Thy-1⁺ cells) were sorted. Whole-cell lysates prepared from the sorted cells were used in Western blots to determine the levels of c-Myc protein. After induction of pre-TCR signaling (24 hours), TetOβ-LTH-Rag1^-/- cells showed an accumulation of c-Myc protein compared with LTH-Rag1^-/- control cells (Figure 1B), indicating that c-Myc was rapidly activated in response to pre-TCR signals. Similar blots detected no changes in the levels of cyclins D2 and D3 24 hours after induction of pre-TCR signaling (data not shown).

c-Myc expression was further examined after induction of pre-TCR like signaling in Rag-deficient mice by α-CD3ϵ treatment. Injection of the α-CD3ϵ mAb (2C11) in Rag-deficient mice has been previously shown to induce pre-TCR like signaling and promote developmental progression of DN3 cells to the DP stage.^50,51  To determine the effect of these signals on c-Myc expression, lysates were prepared from thymocytes isolated 0, 6, 16, and 48 hours after intraperitoneal injection of purified α-CD3ϵ mAb (50 μg/mouse) in Rag2^-/- mice and subjected to Western blot analyses. c-Myc protein levels increased gradually throughout the period of observation (48 hours), starting at 6 hours after α-CD3ϵ injection (Figure 1D). To correlate the c-Myc protein levels with the course of the developmental progression induced by α-CD3ϵ treatment, we stained thymocyte suspensions of similarly treated mice with antibodies against CD4 and CD8, or against lineage markers combined with CD44 and CD25, followed by FACS analysis (Figure 1C). After α-CD3ϵ injection (48 hours), thymocytes had not yet developed to the DP stage and were still undergoing the transition from the DN3 to the DN4 stage. Thus, c-Myc induction preceded developmental progression, indicating that it was a consequence of pre-TCR signaling and not a result of the developmental transition.

Taken together, these data suggest that c-Myc expression is induced by pre-TCR signaling and emphasize the need for detailed analyses to determine its role at the pre-TCR-dependent stages of thymocyte development.

To characterize the role of c-Myc specifically downstream from the pre-TCR, avoiding earlier developmental defects, we used a novel mouse model that allows conditional ablation of c-Myc starting at the DN3 stage of thymocyte development. This was obtained by crossing Myc^fl/fl mice⁴⁶  that carry LoxP sites flanking the coding exons 2 and 3 of the Myc gene with mice expressing Cre under the control of the proximal p56^Lck promoter⁵²  (LckCre).

Cre-mediated deletion of exons 2 and 3 of the Myc gene in compound mutant LckCre-Myc^fl/fl mice was detectable at the DN3 stage (data not shown), and thus ablation of c-Myc was expected to coincide with the onset of pre-TCR signaling. Efficient and stage-specific Cre-mediated ablation of c-Myc in these mice was examined by semiquantitative RT-PCR, using cDNA derived from sorted DN3- and DN4-stage thymocytes. Expression of c-Myc was reduced about 5-fold at the DN3 stage and was completely abrogated at the DN4 stage (Figure 2A). Western blot analyses of extracts from similarly sorted cells showed that LckCre control thymocytes had lower levels of c-Myc protein at the DN3 stage than at the pre-TCR-dependent DN4 stage (Figure 2B). Deletion of the Myc gene in LckCre-Myc^fl/fl thymocytes severely diminished the expression of c-Myc protein both in the DN3 and DN4 subsets.

To determine the impact of c-Myc ablation on thymocyte development we compared the thymocyte subset distribution (Figure 2C) and thymic cellularity (Figure 2D) of LckCre-Myc^fl/fl and control embryos and adult (5- to 8-week-old) mice. Adult LckCre-Myc^fl/fl contained approximately 10 times fewer thymocytes

(1.2 × 10⁷ ± 0.11 × 10⁷) than LckCre controls (8.9 × 10⁷ ± 0.66 × 10⁷). This mainly reflected an approximately 30-fold reduction in the number of CD4⁺CD8⁺ DPs in LckCre-Myc^fl/fl (2.7 × 10⁶ ± 0.17 × 10⁶) compared with the equivalent LckCre control cells (8.4 × 10⁷ ± 0.92 × 10⁷) (Figure 2D), and the reduction of the subsequent SP stages. Cellularity at the DN3 stage of LckCre-Myc^fl/fl mice was elevated compared with LckCre controls (P = .02; Figure 2D). Interestingly, the number of LckCre-Myc^fl/fl DN4-stage thymocytes was comparable with that of LckCre controls, indicating that development through the pre-TCR checkpoint was not inhibited.

Similar results were obtained from embryonic thymi (Figure 2C-D). Control Myc^fl/fl embryonic thymi contained on average 4.2 × 10⁶ ± 1.2 × 10⁶ cells. This number was reduced by approximately 50% in LckCre-Myc^fl/fl mice (2.3 × 10⁶ ± 0.6 × 10⁶), accounting for an approximately 10-fold decrease in the number of DPs (2.2 × 10⁶ ± 0.4 × 10⁶ versus 1.9 × 10⁵ ± 0.8 × 10⁵). The DN compartment remained unaltered in profile (Figure 2C) and cellularity (Figure 2D).

Collectively, these data show that conditional ablation of c-Myc at the pre-TCR checkpoint did not affect the transition to the DN4 stage but resulted in the development of fewer DP cells (19% versus 82% adult, and 8% versus 48% embryonic DP cells; Figure 2C). This could either reflect reduced proliferation or increased apoptosis of cells traversing the pre-TCR checkpoint.

Following assembly of the pre-TCR, developing thymocytes undergo rapid proliferation at the DN4 stage before progressing to the CD4⁺CD8⁺ DP stage. To address the role of c-Myc in this wave of proliferation, we compared the fraction of cycling cells in LckCre control and LckCre-Myc^fl/fl immature thymocytes. Thymocyte suspensions were stained on the surface to allow identification of the various thymocyte subsets, followed by intracellular staining with the DNA-binding dye 7-amino actinomycin D (7-AAD) and FACS analysis. At the resting DN3 stage, only 4% of LckCre and 2% of LckCre-Myc^fl/fl thymocytes were in the G₂/S/M phases of the cell cycle. The fraction of LckCre control cells in the G₂/S/M phases at the actively proliferating DN4 stage was 29.1% ± 1.96%, while only 12.2% ± 0.78% of the c-Myc-deficient thymocytes were cycling (Figure 3A), suggesting that c-Myc was required for proliferation at the pre-TCR checkpoint. We also observed that LckCre-Myc^fl/fl DN4 stage cells were smaller (Figure 3B) than the equivalent LckCre cells, probably reflecting the contribution of c-Myc to cell growth.

To ensure that the ablation of c-Myc did not impact pre-TCR assembly we compared the expression of components of the pre-TCR complex in LckCre control and LckCre-Myc^fl/fl thymocytes. Semiquantitative RT-PCR using RNA prepared from sorted DN3- and DN4-stage thymocytes indicated that the expression of pTα and TCRβ mRNA was not impaired in LckCre-Myc^fl/fl mice. These cells expressed even higher levels of pTα mRNA compared with the equivalent LckCre control cells (Figure 3C). Intracellular TCRβ staining revealed that an equal fraction of DN4-stage thymocytes from LckCre-Myc^fl/fl and LckCre mice expressed TCRβ chains, indicating that c-Myc deficiency did not affect the rearrangement and/or the synthesis of TCRβ chains (Figure 3D). These findings indicate that the c-Myc deficiency affects thymocyte proliferation downstream of a properly assembled pre-TCR.

Although several genes involved in cell-cycle progression and growth control were shown to be transcriptionally regulated by c-Myc,^53,54  the mechanism by which c-Myc mediates cell-cycle progression remains unclear, especially in thymocytes. To trace the impact of c-Myc ablation on genes involved in proliferation, we compared their expression in LckCre-Myc^fl/fl and LckCre DN3- and DN4-stage thymocytes by semiquantitative and quantitative RT-PCR as well as Western blots (Figure 4A-C). These analyses revealed strikingly elevated protein levels of the cell-cycle inhibitor p27^Kip in LckCre-Myc^fl/fl DN3 and DN4 thymocytes (Figure 4C). c-Myc-deficient DN4-stage cells also showed elevated expression of the growth arrest and DNA-damage-inducible factor 45 alpha (Gadd45α), and to a lesser extent, of the cell-cycle inhibitor p21^Cip1 (Figure 4B). Both message and protein levels of the growth-promoting cyclins D2 and D3 appeared unchanged, and the mRNA levels of cyclin E1 were only modestly increased (Figure 4A,C). This is noteworthy considering that cyclin D3 (Figure 4C) has previously been reported to control the proliferative expansion of DN4 and immature single-positive (ISP) thymocytes.⁵⁵  Thus, it is likely that the proliferation defect following c-Myc ablation is related to the elevated expression of cell-cycle inhibitors such as p27^Kip, p21^Cip1, and Gadd45α, which are normally negatively controlled by c-Myc.⁵⁶  This notion was further supported by the unchanged expression of genes previously implicated in thymocyte proliferation. These included the inhibitory protein Tis21, described to regulate stage-specific proliferation in fetal thymocytes,⁵⁷  the Id3 inhibitor of E2A activity¹⁸  previously shown to be induced after pre-TCR signaling⁵⁸  (modestly elevated at the DN4 stage at approximately 70% of control levels; Figure 4B). c-Myc ablation was linked to a probably compensatory transcriptional up-regulation of N-Myc at the DN3 and DN4 stages.

These observations indicate that c-Myc impacts proliferation at the pre-TCR checkpoint by affecting the expression of cell-cycle inhibitors, especially p27^Kip and Gadd45α, rather than directly affecting the expression of the cell-cycle-promoting cyclins or other genes reported to promote thymocyte expansion downstream of the pre-TCR.

LckCre-Myc^fl/fl thymocytes developed to the CD4⁺CD8⁺ DP stage despite reduced proliferation at the DN4 stage, indicating that c-Myc ablation did not influence their differentiation potential. To precisely address the differentiation potential of c-Myc-deficient immature thymocytes we crossed LckCre-Myc^fl/fl mice onto the Rag2^-/- background. These mice showed a complete block at the DN3 stage of thymocyte development. We then induced pre-TCR-like signaling in LckCre-Myc^fl/fl-Rag2^-/- and control LckCre-Rag2^-/- mice by injecting α-CD3ϵ mAb (50 μg/mouse) and analyzed the thymic development 4 days later with respect to cellularity and thymocyte subset distribution. Developmental progression was examined by staining for surface expression of CD4, CD8, CD25, and CD44 (Figure 5A). Both LckCre-Myc^fl/fl-Rag2^-/- and LckCre-Rag2^-/- mice progressed developmentally in response to the treatment; however, LckCre-Myc^fl/fl-Rag2^-/- thymi contained 3- to 5-fold fewer thymocytes than LckCre-Rag2^-/- controls. The 2 mouse strains had comparable numbers of DN3- and DN4-stage thymocytes, indicating that c-Myc ablation did not inhibit progression to the DN4 stage. The reduced cellularity of LckCre-Myc^fl/fl-Rag2^-/- thymi compared with LckCre-Rag2^-/- controls was entirely reflected in the reduced fraction (25% versus 75%) and number (4.0 × 10⁷ ± 7.9 × 10⁶ versus 6.9 × 10⁶ ± 1.73 × 10⁶; Figure 5B) of DP cells. These data suggested that c-Myc was not required for the developmental transition from the DN to DP stage following pre-TCR signaling.

To rule out the possibility that the progressing cells in α-CD3ϵ-treated LckCre-Myc^fl/fl-Rag2^-/- mice may have “escaped” timely Cre-mediated deletion of Myc, we performed semiquantitative PCR analyses using genomic DNA isolated from sorted DN3, DN4, and DP cells. The floxed Myc allele was barely detectable in DN4 and DP stage LckCre-Myc^fl/fl-Rag2^-/- thymocytes, indicating that these cells had undergone efficient Myc deletion (Figure 5C). Moreover, the LckCre-Myc^fl/fl-Rag2^-/- DN4 thymocytes were smaller than the LckCre-Rag2^-/- (Figure 5D), likewise indicating efficient c-Myc ablation. Thus, LckCre-Myc^fl/fl-Rag2^-/- thymocytes developed to the DP stage despite the lack of c-Myc.

To determine whether developmental progression required cell division we cocultured immature LckCre-Myc^fl/fl and LckCre control thymocytes with OP9-DL1⁸  cells. Independently sorted DN3- and DN4-stage thymocytes from LckCre control and LckCre-Myc^fl/fl mice were labeled with CFSE and cocultured with OP9-DL1 cells for 4 days before staining for CD4 and CD8 surface expression and FACS analysis. A substantial fraction of c-Myc-deficient DN3-stage thymocytes up-regulated CD4 (Figure 6A) and CD8 (Figure 6B) surface expression without any cell division, while the developing fraction of LckCre control DN3 cells had undergone 4 to 5 cell divisions. Likewise, c-Myc-deficient DN4 cells did not divide, but up-regulated CD4 and CD8 surface expression. To determine that these DP cells had undergone proper differentiation we examined their surface expression of TCRαβ (Figure 6C). LckCre-Myc^fl/fl and LckCre control DP cells developing in OP9-DL1 cocultures expressed comparable levels of TCRαβ on their surface. These findings provided both in vivo and in vitro evidence that neither c-Myc signaling nor proliferation was required for developmental progression at the pre-TCR checkpoint.

Immature thymocytes that do not receive pre-TCR signals undergo apoptosis. c-Myc has been reported to control cell survival in other experimental systems. To examine whether c-Myc deficiency had an effect on the survival of developing thymocytes, we compared LckCre-Myc^fl/fl and LckCre control thymocytes with respect to their fraction of Annexin V⁺ cells, as well as the expression levels of the antiapoptotic proteins Bcl-2 and Bcl-x_L and the proapoptotic protein p53. To this aim we stained primary thymocytes with Annexin V as well as antibodies directed against surface markers that allow electronic gating of specific thymocyte subsets (Figure 7A). LckCre-Myc^fl/fl and LckCre control mice had comparable fractions of Annexin V⁺ cells in the DN3, DN4, and DP subsets, indicating that the c-Myc deficiency did not affect the survival of developing thymocytes. Intracellular staining with antibodies against Bcl-2 revealed that LckCre-Myc^fl/fl and LckCre thymocytes expressed comparable levels of Bcl-2 at the DN3, DN4, and DP stages, further supporting this notion (Figure 7B). We also analyzed the expression levels of Bcl-x_L and p53 mRNA using semiquantitative RT-PCR (Figure 7C) and of p53 protein using Western blot (Figure 7D). The expression level of p53 was unchanged, while Bcl-x_L expression was modestly elevated in the absence of c-Myc.

In summary, these data indicate that c-Myc ablation at the DN3 stage did not impair the pre-TCR-dependent survival signals, and that the reduced thymic cellularity was entirely the result of reduced proliferation.

Pre-TCR assembly and signaling promotes proliferation, survival, and differentiation of immature thymocytes at the DN3 stage of development, essentially instructing them to the αβT-cell lineage.⁹  Using 2 inducible ways to promote this developmental transition we found that an early event following the onset of pre-TCR signaling was the up-regulation of c-Myc. Thus, within hours following induction of pre-TCR signaling, c-Myc protein levels increased, indicating that this molecule was likely involved in the proliferation, survival, or differentiation processes mediated by pre-TCR signaling. We showed that conditional thymocyte-specific ablation of c-Myc impaired cell growth and proliferation of immature thymocytes at the pre-TCR checkpoint. Despite reduced proliferation, the pre-TCR could still signal differentiation and survival to c-Myc deficient thymocytes both in vivo and in vitro. Our findings provide a dissection of pre-TCR signaling, and assign c-Myc specifically downstream of the proliferation but not the differentiation or survival signals.

Three lines of evidence support the suggestion that c-Myc is dispensable for the differentiation signals downstream from the pre-TCR. First, c-Myc-deficient thymocytes progressed efficiently through the DN4 stage, although they yielded a reduced number of DP cells. Second, induction of pre-TCR-like signaling by α-CD3ϵ treatment of LckCre-Myc^fl/fl-Rag2^-/- mice promoted progression to the DP stage despite the c-Myc deficiency. Third, while control LckCre DN3-stage thymocytes cocultured with OP9-DL1 cells progressed to the DP stage while undergoing an average of 5 cell divisions, a substantial fraction of LckCre-Myc^fl/fl DN3 cells progressed to the DP stage and acquired TCRαβ surface expression with 0 to 1 cell divisions. These data demonstrate that the failure of c-Myc-deficient thymocytes to proliferate does not impact their pre-TCR-dependent differentiation potential. Our observations are in contrast with an earlier report by Douglas and colleagues⁴³  showing that c-Myc-deficient thymocytes are unable to differentiate to the DP and SP stages in Rag^-/-Myc^-/- chimeras. This apparent discrepancy may be related to differences in the experimental systems. While Douglas and colleagues mainly focused on embryonic thymocytes with a constitutive c-Myc deficiency, we conditionally ablated c-Myc immediately prior to the pre-TCR checkpoint, thus avoiding the accumulation of defects from earlier developmental stages.

It is important to distinguish here between the ability of thymocytes to developmentally progress and the number of cells detected in each developmental stage. Since c-Myc-deficient thymocytes progressed to the DP stage without dividing while the equivalent LckCre cells reached this stage after undergoing 4 to 5 cell divisions, an approximately 32-fold reduction in the number of DP cells would be expected in LckCre-Myc^fl/fl mice. This prediction is in line with the reduction in the number of DP cells in LckCre-Myc^fl/fl mice. Similarly, fewer LckCre-Myc^fl/fl-Rag2^-/- thymocytes treated with α-CD3 are expected to reach the DP stage due to the impairment in proliferation. Indeed, we detected a 5-fold reduction in the number of DP thymocytes in α-CD3ϵ-treated LckCre-Myc^fl/fl-Rag2^-/- mice compared with Rag2^-/- controls 4 days after injection. Our findings that c-Myc deficiency does not impair differentiation are in line with recent reports showing that in vitro-cultured, c-Myc-deficient HSCs are able to differentiate along the myeloid and lymphoid lineages.⁴⁵  Interestingly, in the case of the HSCs—like in thymocytes—this occurs in the absence of significant proliferation. Although our data show that in the absence of c-Myc differentiation is in principle not inhibited in vivo, the reduced numbers of terminally differentiated cells may be devastating and even give the impression of a developmental block.

The c-Myc-deficient DN4-stage thymocytes were cycling at lower frequencies and were smaller, indicating that the proliferative signals attributed to the pre-TCR were impaired in the absence of c-Myc. This proliferative block was detected both in vivo and in vitro but was more dramatic in OP9-DL1 stromal cell cocultures seeded with sorted thymocytes from either the DN3 or the DN4 stage. c-Myc-deficient thymocytes failed to undergo more than 1 cell division in these cultures. The mechanism by which c-Myc promotes thymocyte proliferation may rely on the regulation of cell-cycle inhibitors such as p27^Kip, Gadd45α, and p21^Cip1.^59-63  This is indicated by the elevated expression of these molecules in c-Myc-deficient thymocytes. Elevated levels of p27^Kip have been previously proposed to result in proliferation defects of c-Myc-deficient B cells.⁴⁶  Gadd45α has been suggested to determine the susceptibility of a cell to p21^Cip1-induced cell-cycle arrest,⁶⁴  has been linked to T-cell proliferation,⁶⁵  and has also been reported to suppress cell growth via inhibition of the G₂-M-promoting Cdc2 kinase.⁶⁶ 

c-Myc has been shown to promote cell growth in B cells^33,37  and thymocytes.³⁴  By contrast, Trumpp and colleagues⁶⁷  observed that reduced levels of c-Myc do not affect the size of T cells upon activation. Our finding that c-Myc ablation at the DN3 stage of thymocyte development resulted in small cells that failed to develop into blasts at the DN4 stage suggests that while reduced levels of c-Myc may still be sufficient to mediate cell growth, a complete ablation is not. This explanation is in line with recent findings that c-Myc is likely to regulate cellular growth through ribosome biogenesis,^27,68,69  and that this regulation requires only low levels of c-Myc in the nucleolus.

The reduced thymic cellularity observed in LckCre-Myc^fl/fl mice is most likely the result of impaired proliferation, and is not associated with reduced survival. c-Myc has long been thought to sensitize cells to apoptosis, particularly when it is overexpressed. However, conditional ablation of c-Myc was shown not to affect the survival of HSCs⁴⁵  or primary B lymphocytes.⁷⁰  In line with these observations, we found that spontaneous apoptosis of immature c-Myc-deficient thymocytes was comparable with that of control thymocytes at the equivalent stages. Moreover, we did not detect deregulated expression of several genes implicated in cell survival/death that have been classified as potential c-Myc target genes, such as p53, Bcl-2, and Bcl-x_L.^71-73 

Here we show that c-Myc is rapidly induced upon activation of the pre-TCR. This could be directly controlled by pre-TCR signals, or it could be an indirect consequence. However, the rapid up-regulation of c-Myc protein levels within 6 hours after induction of the pre-TCR argues in favor of a direct control of c-Myc by the pre-TCR. Irrespective of the mechanism by which pre-TCR induces c-Myc, our study reveals for the first time a bifurcation of signaling pathways at the pre-TCR checkpoint and shows that differentiation of thymocytes occurs efficiently in the absence of c-Myc-dependent proliferation.

The authors thank R. Chang for excellent technical assistance. A. Parmelee and S. Kwok at the Tufts Laser Cytometry facility provided invaluable help with cell sorting. We thank Dr F. Alt for providing the Myc^fl/fl mice, Dr P. Sicinski for critical reading of the manuscript, and Dr A. Garbe for helpful discussions.