Early Growth Response Gene 3 Regulates Thymocyte Proliferation during the Transition from CD4⁻CD8⁻ to CD4⁺CD8⁺¹

The earliest T cell precursors entering the thymus from the bone marrow do not express CD4, CD8, or the TCR. These early T cell precursors can be subdivided on the basis of CD44 and CD25 expression into the double negative (DN)³ stages I through IV (1). DNI cells are CD44⁺, CD25⁻, and the small population of DNII cells are CD44⁺CD25⁺. For those cells in the αβ T cell lineage, it is at the DNIII stage (CD44⁻CD25⁺) that TCRβ rearrangement occurs (1). A functional TCRβ-chain resulting from successful rearrangement assembles with the pre-TCRα-chain (pTα) and forms the pre-TCR complex on the cell surface. For continued development in the αβ lineage, a successful TCRβ rearrangement and expression of the pre-TCR are required (2). Pre-TCR signals allow survival and further differentiation to the DNIV stage (CD44⁻CD25⁻). Thus, the transition from DNIII to DNIV serves as a developmental checkpoint (referred to as β-selection) that eliminates those cells that are unable to produce a functional TCR β-chain. Numerous cellular events occur at this transition, including down-regulation of CD25, several rounds of proliferation, cessation of rearrangement at the TCR β locus, reduced expression of the Rag1 and Rag2 genes, decreased pTα expression, initiation of transcription at the TCRα locus, and the induction of CD4 and CD8 expression (3).

Changes in the expression of cell surface molecules on developing thymocytes have been well defined, but the molecular basis underlying the cellular events in thymocyte differentiation has not been resolved. For example, how pre-TCR signals induce changes in gene expression that result in all of the differentiation associated with the transition from DNIII to DNIV is largely unknown. Pre-TCR signals are associated with activation of lck, ras, NF-κB, and NFAT (4, 5, 6). In addition, the early growth response (Egr) genes are more highly expressed in the DNIV subset compared with the other DN subsets (7, 8). The Egr genes code for a family of four transcription factors that all contain a zinc-finger DNA binding domain. Egr proteins are rapidly induced by extracellular stimuli in a variety of cell types and are important in translating extracellular signals into changes in gene expression. Egr1, 2, and 3 are all expressed in the thymus, and their expression can be increased by TCR signals (9). Retroviral transduction of Egr1, 2, or 3 into day 14 fetal thymocytes isolated from CD3γ-deficient mice (which are blocked at the DNIII stage) allows some differentiation to the DNIV stage (7). In addition, individual Egr proteins are able to differentially regulate expression of the Rag, Cα, and pTα genes. Thus, it is thought that the Egr proteins each regulate specific aspects of differentiation beyond the β selection checkpoint.

The specific role played by Egr1 in the DN to double positive (DP) transition has been examined in some detail. Overexpression of Egr1 in Rag-deficient thymocytes promotes the differentiation of DN cells into immature CD8 single positive (ISP) cells (8). However, Egr1-deficient mice have a normal distribution of DN subsets, and no defect in differentiation to the DP stage (10). Taken together, these results suggest that Egr1 can probably activate transcription of some of the genes involved in transition to DP but that it does not have a nonredundant function. Although Egr2 and 3 have also been implicated in β-selection (7), the specific roles played by these factors in vivo has not been determined.

In the current study, we have evaluated the specific role played by Egr3 in DN thymocytes using Egr3-deficient mice and mice that constitutively express Egr3 in thymocytes. Egr3-deficient mice have impaired muscle spindle development and suffer from sensory ataxia (11). Some Egr3-deficient mice die shortly after birth, but most survive through adulthood and are available for analysis. Thymuses from Egr3-deficient mice have low cellularity that is a result of poor proliferation subsequent to pre-TCR signals. However, many of the differentiation events that occur subsequent to pre-TCR signaling are normal in the absence of Egr3. Analysis of mice that express high levels of transgenic Egr3 has revealed that overexpression of Egr3 in the thymus results in increased cell cycle activity, but a small thymus due to impaired survival. Finally, overexpression of Egr3 in Rag1-deficient thymocytes promotes some aspects of thymocyte differentiation beyond the β-selection checkpoint.

Egr3-deficient mice were kindly provided by Jeffrey Milbrandt and were crossed to and maintained on a B6.AKR background. B6.AKR mice were purchased from The Jackson Laboratory (Bar Harbor, ME). To generate Egr3 transgenic mice, the Egr3 coding region was synthesized by PCR from a plasmid containing the rat Egr3 cDNA using the following primers: 5′-CCCCGGATCCATGACCGGCAAACTCGCCGA, 5′-CCCCCTCGAGTCAGGCGCAGGTGGTGACC. The PCR products were cloned into the BamHI/XhoI site of the pCMV-Tag2 vector (Stratagene, La Jolla, CA) to generate an N-terminal FLAG-tagged Egr3. The full-length FLAG-Egr3 was PCR amplified using the following primers: 5′-GCCCGAATTCAATTAACCCTCACTAAAGGG, 5′-GCCCGAATTCTAATACGACTCACTATAGGG and subcloned into the EcoRI site of the VA-hCD2 vector, which contains a CD2 promoter and a locus control region (12). A 12.55-kb SalI-XbaI fragment containing the CD2 promoter, FLAG-Egr3, and locus control region was injected into pronuclei of fertilized oocytes from C57BL/6 and FVB/N mice by the Emory University Transgenic Mouse Core Facility. Transgenic founders were detected by Southern blot analysis of genomic tail DNA using the full-length rat Egr3 cDNA as a probe. The transgenic lines were propagated by sequential crossing to B6.AKR mice. Two transgenic founder lines, Tg16 and Tg96, both expressing high levels of the FLAG-Egr3 transgene and displaying similar phenotypes, were derived from C57BL/6 and FVB/N mice respectively. The Rag1-deficient mice were purchased from The Jackson Laboratory.

The following Abs for flow cytometry were obtained from BD PharMingen (San Diego, CA): anti-CD4-PE (GK1.5), anti-CD5-FITC (53-7.3), anti-TCRβ-biotin (H57), anti-Gr-1-FITC (RB6-8C5), anti-γδ TCR-FITC (GL3), anti-B220-FITC, anti-CD11b-FITC (M1/70.15), anti-HSA-FITC (M1/69), anti-CD25-FITC (7D4), and anti-CD44-biotin (IM7). The following Abs and reagents were purchased from Caltag (Burlingame, CA): anti-CD8α-TC (Ly-2), anti-CD8β-PE (Ly-3), anti-CD3-FITC (CD3ε), anti-CD25-PE (PC61 5.3), anti-CD4-APC (RM4-5), anti-CD8α-APC (CT-CD8a), streptavidin-tricolor (TC), and streptavidin-FITC. For in vivo 5-bromo-2′-deoxyuridine (BrdU) labeling experiments, mice were injected with BrdU (BD PharMingen) (70 μg/g of body mass, i.p.). Five hours after injection, thymocytes were isolated and surface stained with anti-CD4-PE and anti-CD8-TC or anti-CD4-APC, anti-CD8α-APC, anti-CD25-PE, and anti-CD44-TC. The cells were then fixed and intracellularly stained with anti-BrdU-FITC using the BrdU flow kit from BD PharMingen. For cell cycle analysis, 3 × 10⁶ freshly isolated thymocytes were fixed in 200 μl 70% ethanol at 4°C for 1 h, washed with PBS, and stained with 10 μg/ml 7-amino-actinomycin D (BD PharMingen) in 100 μl PBS at 4°C for 30 min. Thymocyte apoptosis was analyzed using the annexin V-FITC apoptosis detection kit I (BD PharMingen) according to the manufacturer’s instructions. For the in vitro survival assay, thymocytes (3 × 10⁶ per well in a 24-well plate) were cultured for various periods of time and analyzed using the annexin V apoptosis detection kit. All cell analyses were performed with a FACSort flow cytometer and CellQuest software (BD Immunocytometry Systems, San Jose, CA).

Thymic Cdk2 kinase activity was analyzed using the cdk1/cdc2 kinase assay kit (Upstate, Charlottesville, VA) with minor modifications. Briefly, 500 μg of thymocyte whole-cell lysate was immnuoprecipitated with rabbit anti-Cdk2. Immune complexes were washed three times with lysis buffer and three times with kinase buffer (10 mM Tris, pH 7.4, 150 mM NaCl, 10 mM MgCl₂, and 0.5 mM DTT). Immune complexes were then incubated with 50 μl of kinase buffer containing 20 μg histone H1, 10 μCi [γ-³²P]ATP and protein kinase C/protein kinase A inhibitor cocktail at 30°C for 10 min. The reaction was terminated by incubation at 100°C for 5 min. The reaction mixture was separated on a 12% SDS-PAGE gel, which was then stained with Ponceau S to locate the histone H1 protein. The gel was dried and autoradiographed. For quantification of histone phosphorylation, the gel slice containing the histone band was excised and subjected to scintillation counting. Immunoblotting of the immunoprecipitated Cdk2 was used to normalize the amount of input Cdk2.

Freshly isolated thymocytes were lysed with the lysis buffer containing 150 mM NaCl, 10 mM Tris-HCl, pH 7.6, 5 mM EDTA, 1% Nonidet P-40, 1 mM Na₃VO₄, 10 μg/ml leupeptin, 10 μg/ml aprotinin, and 1 mM phenolmethylsulfate). The lysates were incubated on ice for 15 min and then centrifuged at 12,000 rpm for 10 min at 4°C. Protein concentration of the supernatant was determined by the Bradford assay (Pierce Biotechnology, Rockford, IL). The whole-cell lysates (50 μg protein) were subjected to SDS-PAGE, and the separated proteins were transferred to a nitrocellulose membrane (Amersham Biosciences, Piscataway, NJ). The following Abs were used as probes: HRP-anti-FLAG M2 mAb (Sigma-Aldrich, St. Louis, MO), anti-Egr3 Ab (C-24; Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anti-p27 Ab (C-19; Santa Cruz Biotechnology), anti-β-actin (AC-15; Sigma-Aldrich). Immune complexes were detected with appropriate HRP-conjugated Abs (Jackson ImmunoResearch, West Grove, PA) and ECL reagents (Amersham Biosciences).

Total RNA was isolated from freshly isolated thymocytes using the TRIzol reagent (Invitrogen, Carlsbad, CA) according to manufacturer’s instructions. cDNA was synthesized with Superscript II reverse transcriptase (Invitrogen) with random hexamers. The expression of TCR-Cα, pTα, and hypoxanthine phosphoribosyltransferase (HPRT) was analyzed by semiquantitative PCR with serially diluted cDNA (1/5) as follows: 94°C for 2 min and then 35 cycles of 94°C for 1 min, 54°C for 1 min, and 72°C for 1 min, followed by 72°C for 10 min. The PCR primers for Cα and pTα were as described (13). The PCR primers for HPRT are as follows, 5′-GCTGGTGAAAAGGACCTCT, 5′-CACAGGACTAGAACACCTGC. PCR conditions for endogenous Egr3 and FLAG-Egr3 were as follows: 94°C for 2 min and then 30 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 1 min, followed by 72°C for 10 min. PCR primers for FLAG-Egr3 are as follows: 5′-GGATTACAAGGATGACGACGAT, 5′-GTCGTTGGGGTGGTGATACAGG. PCR primers for endogenous Egr3 are as follows: 5′-CGACTCGGTAGCCCATTACAATCAGA, 5′-GAGATCGCCGCAGTTGGAATAAGGAG. PCR products were separated on a 1.5% agarose gel and transferred to a Hybond N nylon membrane (Amersham Biosciences) for analysis by Southern blot. Probes for Egr3, Cα, and pTα were full-length cDNAs, and the probe for HPRT was the following oligonucleotide: 5′-TACGAGGAGTCCTGTTGATGTTGCCA.

Published data has demonstrated that Egr3 is more highly expressed in the DNIV subset than in the DNIII subset, suggesting that Egr3 expression is induced by pre-TCR signaling (7). We have examined this issue by using RT-PCR to detect Egr3 mRNA in thymocytes from Rag1-deficient mice injected with either PBS or anti-CD3ε Ab. Rag1-deficient thymocytes stop development at the DNIII stage, and anti-CD3ε cross-linking on these thymocytes mimics pre-TCR signaling and induces development to the DNIV stage and beyond (14). Rag1-deficient thymocytes injected with PBS have barely detectable amounts of Egr3 mRNA, but anti-CD3ε injection induces high amounts of Egr3 message that is sustained for at least 8 h (Fig. 1). These results suggest that pre-TCR signals in DNIII thymocytes rapidly induce high levels of Egr3 expression.

The high levels of Egr3 induced in thymocytes in the transition from DN to DP raises the possibility that Egr3 may control specific aspects of this differentiation step. As a first approach to address this question, we analyzed thymocytes in Egr3-deficient mice. As shown in Fig. 2,A, thymuses from Egr3-deficient mice had only 41% of the number of cells found in age-matched wild-type controls. CD4 and CD8 staining revealed that although these thymuses were smaller, the distribution of cells among DN, DP, and single positive (SP) compartments was mostly normal, with a slight increase in the percentage of DN cells (Fig. 2, B and C). A closer look at the DN compartment showed that there was some impairment in the transition from the DNIII to the DNIV stage, with Egr3-deficient thymuses having a higher percentage of the DN cells at the DNIII stage compared with wild-type controls (Fig. 2, D and E). This made the ratio of DNIV to DNIII cells in Egr3^−/− mice (0.43) to be only about half as large as the ratio in Egr3^+/+ mice (0.83). Thus, Egr3-deficient thymuses either have a defect in differentiation past the DNIII stage, poor proliferation of DNIV cells, or poor survival of cells that have passed the β-selection checkpoint.

Any one of these three possibilities (impaired differentiation, proliferation, or survival) could account for the overall low cellularity of Egr3-deficient thymuses. We examined thymocyte proliferation by injecting BrdU into Egr3^−/− and Egr3^+/+ mice. Five hours after injection, thymocytes were removed, stained with Abs against CD4, CD8, and BrdU, and the percentage of cells in each subset that had incorporated BrdU was determined. The data shown in Fig. 3,A indicate that Egr3^−/− thymocytes have significantly reduced proliferation when compared with Egr3^+/+ mice. A total of 14.8% of the thymocytes incorporated BrdU in Egr3^+/+ mice vs only 7.7% in Egr3^−/− (p < 0.005 by Student’s t test). Importantly, the reduced proliferation was observed in the DN and CD8SP subsets that contain the highest proportion of cells proliferating in response to pre-TCR signals. Gating on the different DN subsets revealed that Egr3^−/− DNIII and DNIV cells also had about a 50% reduction in BrdU uptake, whereas DNI cells from Egr3^−/− mice had BrdU incorporation similar to Egr3^+/+ mice. Survival was also examined by culturing thymocytes from Egr3^−/− and Egr3^+/+ mice and calculating the percentage of live cells remaining at various time points over 48 h. This experiment showed that survival of Egr3^−/− thymocytes was not impaired in vitro (Fig. 3,B). In addition, we determined the percentage of apoptotic cells in Egr3^−/− thymuses by annexin V and propidium iodide staining and found that it was not significantly different from wild-type controls (Fig. 3 C). These data are consistent with the low cellularity of Egr3^−/− thymuses being a result of either a defect in proliferation or an inability to differentiate past the DNIII stage.

To look at specific differentiation events occurring subsequent to pre-TCR signaling, we bred the Egr3^−/− mice to Rag1^−/− mice. Almost all of the thymocytes in Rag1^−/− mice are at the DNIII stage, and treatment of Rag1^−/− thymocytes with anti-CD3ε mimics pre-TCR signaling and induces the events associated with the transition past the β-selection checkpoint (14). Therefore, we compared cellular and molecular events occurring in response to anti-CD3ε treatment in Egr3^−/−, Rag1^−/− doubly deficient mice to similarly treated Egr3^+/+, Rag1^−/− mice. As shown in Fig. 4,A, Egr3^+/+ and ^−/− mice on the Rag1 deficient background had similar numbers of thymocytes, and a similar distribution of cells among the four DN subsets (95–98% of cells at the DNIII stage). Treatment with anti-CD3ε resulted in almost complete down-regulation of CD25 in both Egr3^−/− and Egr3^+/+ thymocytes after 4 days. CD4 and CD8 were also induced to a similar degree in Egr3^−/− and Egr3^+/+ thymocytes, producing a population of DP thymocytes (69 and 63% respectively). In addition, Cα transcription was induced by ant-CD3ε in both Egr3^−/− and Egr3^+/+ mice, and levels of pTα mRNA were markedly reduced. The striking difference between Egr3 ^−/− and Egr3^+/+ mice in these experiments was the cell recovery. On average, Egr3^+/+ thymocytes expanded from 2 million cells to 33 million cells in 4 days, whereas Egr3-deficient thymocytes expanded from ∼2 million cells to only 7.7 million. The cell recovery of anti-CD3ε-treated Egr3^−/−, Rag1^−/− mice was therefore only 23% as much as similarly treated Egr3^+/+, Rag1^−/− mice. In addition, cell cycle analysis showed that a lower percentage of cells from Egr3^−/−, Rag1^−/− mice were in the S/G₂/M phase of the cell cycle (Fig. 4 C), suggesting that enhanced death of Egr3^−/− cells did not play a role in the poor cell recovery. These results suggest that Egr3-deficient mice have impaired proliferation in response to pre-TCR signaling, but differentiation past the DNIII stage (CD25 down-regulation, CD4 and CD8 expression, induction of Cα transcripts, and pre-Tα down-regulation) is normal.

The analysis of Egr3-deficient mice demonstrated that loss of Egr3 results in poor proliferation during the DN to DP transition, with little effect on differentiation that occurs at this transition. Because Egr3 is part of a family of transcription factors, and there are examples of functional redundancy between Egr3 and Egr2, we also wanted to examine how constitutive overexpression of Egr3 influenced thymocyte development. Toward this end, we generated transgenic mice that expressed a FLAG-tagged version of the rat Egr3 cDNA under the control of the CD2 promoter and enhancer (Egr3TG). These elements are known to drive expression in thymocytes and mature T cells. Two founder lines were generated (Tg16 and Tg96) that had very similar phenotypes, and thymuses from each of these lines were analyzed by Western blot for expression of the endogenous Egr3 as well as the FLAG-tagged transgenic Egr3 (Fig. 5). Probing with anti-FLAG Ab showed that both mouse lines expressed the transgene, and probing with an anti-Egr3 Ab demonstrated that levels of the Egr3 protein were increased in the transgenic mice well above levels in normal mice. Thus, the Egr3TG mice have high-level expression of Egr3 in thymocytes.

Because Egr3-deficient mice have low numbers of thymocytes, it might be expected that Egr3TG thymuses would have increased cellularity compared with nontransgenic controls. However, the cellularity of Egr3TG thymuses was typically only 25% that of littermate control thymuses. The reason for this is that the Egr3TG mice have a defect in survival of DP thymocytes and increased apoptosis of cells at the DP stage (H Xi and G. J. Kersh, manuscript in preparation). However, when we focused on the DN cells in Egr3TG mice we did find that there was an increase in the ratio of DNIV to DNIII cells compared with nontransgenic controls (Fig. 5, C and D). The ratio of DNIV to DNIII cells in Egr3TG mice was 2.0, whereas it was only 1.12 in littermate controls. This is in contrast to the decrease in this ratio that was observed in Egr3^−/− mice (Fig. 2 E). This result supports the idea that Egr3 promotes proliferation and/or differentiation in response to pre-TCR signals.

Because the analysis of Egr3-deficient mice suggested that Egr3 promotes proliferation in response to pre-TCR signals, we looked for evidence of enhanced cell cycle activity during the DN to DP transition. As shown on Fig. 6,A, thymocytes from Egr3TG mice had a much higher proportion of cells with high forward scatter, suggesting that a higher percentage of cells in Egr3TG mice are in the cell cycle. In addition, thymocytes from Egr3TG mice had decreased levels of p27^kip1 (Fig. 6,B), a negative regulator of entry into the cell cycle whose role is to inhibit the function of cyclin-dependent kinase 2 (CDK2) (15). The low levels of p27^kip1 were correlated with elevated CDK2 activity (Fig. 6,C). Another indication that Egr3TG thymocytes have increased proliferation was the increased incorporation of BrdU by Egr3TG DP thymocytes after a 5-h pulse (Fig. 6 D). Therefore, although the Egr3TG thymuses are smaller due to increased apoptosis, they do have evidence of increased cell cycle activity, once again suggesting that Egr3 can promote proliferation in response to pre-TCR signals.

To more clearly focus on the DN thymocytes in Egr3TG mice, and to try to determine whether the enhanced death of noncycling DP cells resulted in the greater cell cycle activity in Egr3TG mice, these mice were bred to Rag1^−/− mice. As shown in Fig. 7,A, the Egr3 transgene was expressed in thymuses of Egr3TG, Rag1^−/− mice, and Egr3 expression on the Rag1-deficient background induced some aspects of the transition from DN to DP in the absence of pre-TCR signaling. Most notably, Egr3 expression resulted in a mean increase in the cellularity of the Rag1^−/− thymuses of 5.7-fold. This result supports the idea that the increased percentage of cycling cells in Egr3TG mice is a result of enhanced proliferation and not merely due to rapid death of noncycling cells. In addition, Egr3 induced a few cells to express CD8, and based on CD24 (heat stable Ag, HSA) expression, these cells were more mature than their CD8 negative counterparts. The Rag1^−/−, Egr3TG thymocytes also had a significant number of cells that had down-regulated CD25 expression (Fig. 7,C), indicating that expression of Egr3 is capable of supporting some of the differentiation beyond the DNIII stage. In support of this, Rag1^−/−, Egr3TG thymocytes had increased levels of Cα mRNA and decreased levels of pTα (Fig. 7 D).

The developmental transition that is initiated by the pre-TCR includes multiple cellular events. Resting DN cells undergo up to nine rounds of cell division in response to pre-TCR signals (16) and differentiate to become CD4, CD8 DP cells that are susceptible to the processes of positive and negative selection. This differentiation includes reduced expression and function of the Rag genes until proliferation is completed, transcription of the Cα gene as a prelude to TCRα rearrangement, and down-regulation of CD25 (3). The transition also includes migration from the subcapsular zone of the outer cortex toward the center of the cortex in the direction of the medullary regions (17). Another aspect of the DN to DP transition that is regulated by pre-TCR signals is survival. Thymocytes are susceptible to death without the proper signals that induce survival, and the pre-TCR signal contributes to thymocyte survival (18). Thus, the signals emanating from the pre-TCR simultaneously induce survival, migration, differentiation, and proliferation in thymocytes. A molecular understanding of how signals from this single receptor can initiate such diverse processes has not been completed.

Many of the proximal signals initiated by the pre-TCR have been fairly well characterized. Once the TCRβ-chain is successfully rearranged and expressed, it forms a pre-TCR with the CD3 chains and the pTα-chain (19). This pre-TCR spontaneously localizes in glycolipid-enriched microdomains in close proximity to signaling molecules (5), and it apparently initiates signals in a ligand-independent manner (20). The proximal signals initiated by the pre-TCR seem to be very similar to proximal signals emanating from the TCR on mature T cells. Lck or fyn is required to phosphorylate immunoreceptor tyrosine-based activation motifs (21), and Zap-70 or syk plays a vital role in phosphorylating downstream substrates such as LAT and slp-76 (22, 23, 24, 25). Further downstream, these signals result in a biphasic rise in intracellular Ca²⁺ concentration that activates the transcription factors NF-κB and NFAT (6). In addition, the mitogen-activated protein kinase cascade is activated (26). Thus, the transcription factors NF-κB and NFAT are activated by pre-TCR signaling, and expression of Egr3 is highly elevated. It is reasonable to assume that Egr3 expression is controlled by NFAT proteins at the DN stage, as Egr3 induction has been shown to be cyclosporin A sensitive in other systems (9). Data also suggests that Egr1 and Egr2 are induced by pre-TCR signals: Egr1 and Egr2 are found at high levels in the DNIV subset, and their transcription is induced by the mitogen-activated protein kinase pathway and NFAT, respectively (9, 27, 28).

The identification of the Egr genes, NFAT, and NF-κB as transcription factors downstream of the pre-TCR provides a framework for deciphering how multiple complex cellular changes are induced simultaneously by a single receptor. In a previous study, we found that Egr1-deficient mice do not have any changes in the DN to DP transition. However, it has been demonstrated in two separate systems that overexpression of Egr1 on backgrounds that lack a pre-TCR signal can result in some development to the ISP stage (7, 8). Thus, it is possible that Egr1 can regulate some aspects of differentiation in response to pre-TCR signals, but it is likely that its role is redundant to other Egr proteins. Egr2 can induce some development past the DNIII stage when overexpressed on a background that cannot produce a pre-TCR signal (7), but a study of thymocyte development in Egr2-deficient mice has not been reported. In the current study, we have found a clear role for Egr3 in promoting proliferation in response to pre-TCR signals; however, the Egr3-deficient mice do not have a complete block in proliferation subsequent to pre-TCR signaling. The partial proliferative defect in Egr3-deficient thymocytes raises the possibility that other Egr proteins can contribute to induction of proliferation in response to pre-TCR signals. We consider Egr2 as the most likely candidate for this function, as Egr2 and 3 have both been shown to be regulated by NFAT (28), and they have previously been shown to have redundant function in the regulation of Fas ligand (29, 30).

Whether NFAT proteins play any role in pre-TCR signaling beyond the induction of Egr2 and 3 transcription is an open question. NFAT3 is not expressed in the thymus, and mice deficient in NFAT1, NFAT4, or doubly deficient for NFAT1 and NFAT4 appear to be normal at the DN to DP transition and have normal-sized thymuses (31, 32, 33). However, the NFAT1/NFAT4 doubly deficient mice have severely reduced expression of Egr2 and 3 in peripheral T cells, resulting in impaired Fas ligand expression and lymphoproliferation. The lack of a defect in proliferation at the DN to DP transition in NFAT1/NFAT4 double knockouts is probably due to the expression of NFAT2 (NFATc). Thymocytes deficient for NFAT2 have decreased proliferation at the DN to DP transition, suggesting that NFAT2 is more important for regulating proliferation at this stage than the other family members (34, 35). A possible explanation for this is that NFAT2 regulates Egr3 expression at the DN stage, whereas NFAT1 and NFAT4 regulate Egr3 expression in mature T cells. How and why different NFAT family members regulate Egr3 at different stages of development is not known, but it is also possible that NFAT2 has an effect at the DN stage that is independent of Egr3. In contrast to Egr3 and NFAT2, the role of NF-κB in the transition to DP is to promote survival of cells after pre-TCR signaling. Thymuses that have reduced NF-κB activity also have low numbers of DNIV cells even though they have normal percentages of BrdU-positive cells (18).

The picture that emerges is one where each transcription factor that is induced downstream of pre-TCR signals regulates a specific event in the transition to DP thymocyte. In the most simplistic model, NF-κB is induced to regulate survival, and Egr3 is induced to regulate proliferation. Whether or not Egr3 also regulates specific differentiation events is not clear. The data from Egr3^−/− mice suggest that Egr3 only regulates proliferation and other aspects of the transition are normal. The low number of DNIV cells in Egr3^−/− mice is probably due to poor expansion of those cells. However, in the Egr3TG mice on a RAG1^−/− background we did observe CD25 down-regulation and induction of Cα transcripts. This raises the possibility that Egr3 may influence gene expression in addition to inducing proliferation. There may be redundancy in Egr3-deficient mice, and other factors substitute for Egr3 and maintain normal levels of CD25 and Cα expression. Alternatively, Egr3 may be influencing gene expression in the Egr3TG mice because it is over expressed, and normal levels of Egr3 would not affect these genes. A third possibility is that changes in CD25 and Cα expression are a response to proliferation at the DNIV stage. By inducing proliferation, Egr3 would be influencing differentiation, but only in an indirect way. Overall, we can conclude that Egr3 plays a significant role in regulating proliferation in response to pre-TCR signals, and it may play an additional role in the induction of specific genes during the transition to DP.

Because Egr3 is a DNA-binding transcriptional regulator, the expectation is that it controls proliferation in response to pre-TCR signals by regulating the expression of specific target genes that are responsible for cell cycle progression. However, few target genes for Egr3 have been identified, and none of the known targets are good candidates for cell cycle regulation. Although decreasing Egr1 expression has been shown to inhibit proliferation of prostate cancer cells, the current study is, to our knowledge, the first to suggest a role for Egr3 in regulation of proliferation. Thus, the identification of Egr3 target genes that are responsive to pre-TCR signals and regulate proliferation is a goal of future studies.

We thank Jeffrey Milbrandt for the Egr3-deficient mice and Brian Evavold for critical review of the manuscript.