Mastermind-1 is required for Notch signal-dependent steps in lymphocyte development in vivo

Abstract

Mastermind (Mam) is one of the elements of Notch signaling, an ancient system that plays a pivotal role in metazoan development. Genetic analyses in Drosophila and Caenorhabditis elegans have shown Mam to be an essential positive regulator of this signaling pathway in these species. Mam proteins bind to and stabilize the DNA-binding complex of the intracellular domains of Notch and CBF-1, Su(H), Lag-1 (CSL) DNA-binding proteins in the nucleus. Mammals have three Mam proteins, which show remarkable similarities in their functions while having an unusual structural diversity. There have also been recent indications that Mam-1 functionally interacts with other transcription factors including p53 tumor suppressor. We herein describe that Mam-1 deficiency in mice abolishes the development of splenic marginal zone B cells, a subset strictly dependent on Notch2, a CSL protein and Delta1 ligand. Mam-1 deficiency also causes a partially impaired development of early thymocytes, while not affecting the generation of definitive hematopoiesis, processes that are dependent on Notch1. We also demonstrate the transcriptional activation of a target promoter by constitutively active forms of Notch to decrease severalfold in cultured Mam-1-deficient cells. These results indicate that Mam-1 is thus required to some extent for Notch-dependent stages in lymphopoiesis, thus supporting the notion that Mam is an essential component of the canonical Notch pathway in mammals.

Notch signaling is one of the small number of signaling pathways frequently used during the development of metazoans to control different cell fate decisions. It mediates local cell–cell communications by using receptors and ligands that are present on the cell surface (1). In humans, abnormalities in Notch signaling have been linked to a number of diseases. They include T cell acute lymphoblastic leukemia (2) and aortic valve disease (3), which involve the up- and down-modulation of the signaling, respectively.

The Notch genes encode heterodimeric type I transmembrane receptor molecules. On binding with type I transmembrane ligands (Delta and Serrate/Jagged), membrane-bound proteases are activated and thus cleave Notch. As a result, the intracellular (IC) domain of the receptor is released from the membrane. Acting as a second messenger in the signaling pathway, the NotchIC domain is transported to the nucleus, and then participates in transcriptional activation through an association with promoter elements via CBF-1, Su(H), Lag-1 (CSL) [recombination signal sequence-binding protein J (RBP-J) in mammals] DNA-binding proteins. This signaling can be experimentally mimicked by expressing the IC domains of Notch. In mammals, Hes1 is among the primary target genes of the Notch signaling pathway (1, 4).

Several loss-of-function studies have revealed that Notch signaling plays an essential role in the multiple developmental stages for hematopoiesis. Notch1 has been shown to be required for the generation of definitive hematopoiesis (5). A loss of either Notch1 or RBP-J results in the cell-autonomous blockade of T cell lineage development and the emergence of B cells in the thymus, indicating their involvement in T/B lineage commitment (6, 7). A loss of RBP-J but not Notch1 results in an enhanced generation of γδ T cells (8, 9). Both Notch1 and RBP-J have also been shown to be required for the efficient generation of CD4⁺CD8⁺ double-positive (DP) thymocytes from CD4⁻CD8⁻ double-negative (DN) cells in a cell-autonomous fashion (8, 9). In B cell lineage, the development of marginal zone B (MZB) cells in spleen requires both Notch2 and RBP-J in a cell-autonomous manner (10, 11). Delta1 has been shown to be required for the generation of MZB cells in a cell-nonautonomous manner, thus validating it as a ligand for signaling (12). In addition, there is in vitro evidence that Notch signaling suppresses a potential of DN thymocytes to differentiate into natural killer cells (13).

Mastermind (Mam) is one of the elements of Notch signaling that has been evolutionarily conserved from nematodes to humans. Genetic analyses in Drosophila and Caenorhabditis elegans have shown Mam to be an essential positive regulator of this pathway in these species, which apparently possess a single Mam gene in their genome (14, 15). We and others have identified a human Mam family of proteins, consisting of three members (Mam-1, Mam-2, and Mam-3) and thus elucidated their biochemical mechanisms of action (16–19). All of the Mam proteins bind to and stabilize the DNA-binding complex of the NotchIC and CSL proteins in the nucleus. Neither the single NotchIC nor CSL proteins are able to associate with Mam proteins. Concomitant with this ternary complex formation, the activation of transcription from target promoters is potentiated. The three human Mam proteins show remarkable similarities in their functions while having an unusual structural diversity. For the complex formation, three Mam proteins exhibit little preference against four kinds of mammalian Notch moieties. Furthermore, all Mam can augment the transcription evoked by Notch1 and Notch2 to comparable degrees (16, 17). Recently, the crystal structures for the core elements of the Mam–NotchIC–CSL ternary complex bound to DNA have been described (20, 21). These structures are also consistent with the model that we have previously proposed. Furthermore, there have been recent indications that Mam-1 functionally interacts with other transcription factors including p53 tumor suppressor (22, 23).

As first described in Drosophila, Notch signaling can be inhibited by expressing N-terminal basic domains of Mastermind proteins (24), which serves as the binding domains to NotchIC–CSL complex in the native protein (16, 18). In mice, the expression of such truncated forms of Mam-1 (DNMAML1) causes a block in T cell development, the appearance of intrathymic B cells, and a decrease in MZB cells (25, 26). Because such suppression apparently has no specificity, the mutant has been proposed to be a pan-Notch inhibitor.

The significance of the three mammalian Mam species in vivo is still elusive. To elucidate this in vivo function of Mam-1 (also known as MAML1), we have generated a targeted allele of its gene in the mouse. We confirmed a recent report that a deficiency of this gene causes growth retardation and muscular dystrophy (22), while also making an extended analysis regarding its contribution to hematopoiesis. We herein show that Mam-1 is cell-autonomously required for the generation of MZB cells, and the efficient development of early thymocytes, which are both steps that depend on Notch2 and Notch1, respectively. Furthermore, we also show that Mam-1 is dispensable for the generation of definitive hematopoiesis, the generation of the most primitive T lineage progenitors in the thymus, a lineage choice between T/B lymphocytes and the suppression of differentiation into natural killer cells and γδ T cells. These results indicate that Mam-1 is required in part for the Notch signal-dependent stages in hematopoiesis and lymphopoiesis. Furthermore, these results support the notion that Mam is an essential component of Notch signaling in not only in Drosophila and C. elegans but also in mammals.

Results

Targeting of the Mam-1 Gene.

To explore the physiological roles of Mam-1 in mammalian development, we have generated a gene-targeted allele of Mam-1 in murine ES cells. A targeting vector was constructed to delete exon 1 of the Mam-1 gene, which encodes the initiator methionine codon and the basic domain of the protein that is essential for the association with Notch and RBP-J (Fig. 1A) (16). Correct targeting was verified by Southern blot analyses (Fig. 1B). Clones were used to generate chimeric mice, which were mated with C57BL/6 mice. Germline transmission was obtained with one of the clones. Mice heterozygous for the targeted allele were apparently healthy and also showed a normal size.

Fig. 1.

Targeted disruption of the mouse Mam-1 gene. (A) Schematic representations of the wild-type Mam-1 allele around the exon 1, targeting construct, and mutant allele. Probes for the Southern blot analysis are indicated. RI, EcoRI; RV, EcoRV. (B) Southern blot analysis of DNAs isolated from wild-type (+/+), heterozygous (+/−), and homozygous (−/−) mice. The restriction enzymes and probes used are indicated. The sizes of the hybridizing fragments are also indicated.

A primary culture of embryonic fibroblasts (EFs) was obtained from embryonic day 14.5 (E14.5) embryos generated by intercrossing the heterozygous mice. As shown in Fig. 2A, the Mam-1 protein is robustly expressed in the cells of wild-type genotype as assessed by a Western blotting analysis. In contrast, the Mam-1 protein was undetectable from homozygous cells (Fig. 2A). The expression of Mam-1 in the heterozygous cells was between those of wild-type and homozygous cells (Fig. 2A). The expressions of Notch1, Notch2, and Hes1 did not vary in these cells (Fig. 2A). These results indicate that Mam-1 is thus inactivated by gene targeting.

Fig. 2.

The deficiency of Mam-1 results in a marked decrease of transactivation induced by expression of Notch1IC. (A) A Western blot analysis of the extracts obtained from EFs. Genotypes of the cells, mobility of size markers, and identity of the bands are shown. TM, transmembrane subunit; EC, extracellular subunit. (B) Real-time PCR analysis of RNAs isolated from EFs. Genotypes of the cells and mRNA species analyzed are shown. (C) A luciferase assay with EFs. The cells were transfected with a reporter, an internal control for transfection, and expression vectors for indicated proteins (+) or its empty vectors (−). The vertical axis represents the mean value of normalized relative luciferase activity to the mean activity of the wild-type cells transfected with empty vector controls. The error bars indicate SD (n = 3).

We also quantitated the amount of mRNAs for Mam-1, Mam-2, Mam-3, and Hes1 in the fibroblasts by real-time PCR. As shown in Fig. 2B, the amount of Mam-1 mRNA in the heterozygous cells is about one-half of that in the wild-type cells. In homozygous cells, a small but certain amount of the RNA (≈50-fold smaller in amount) was detected by the primers directed against a portion of Mam-1 sequence encoded by the exon 5 (Fig. 2B). The RNA thus detected may represent the species initiated from cryptic promoter and/or unstable mRNA, which is devoid of the sequences encoded by exon 1. The amounts of Mam-2 and Mam-3 mRNAs were not significantly altered by the status of Mam-1 (Fig. 2B). Consistent with the Western blotting findings, the expression of Hes1 did not vary in the three cell types, thus indicating that Notch signaling is “off” in such cultures (Fig. 2B).

Full Activity of Notch Signaling Is Dependent on Mam-1.

To evaluate the contribution of Mam-1 in Notch signaling, a reporter assay was performed with the EFs. As shown in Fig. 2C, a sensitive reporter of intracellular Notch signaling, pTP1-luc was activated by the expression of Notch1IC by ≈100-fold in the wild-type cells. The expression of Mam-1 alone did not significantly activate this promoter. The coexpression of Mam-1 with Notch1IC potentiated the activation ≈2-fold. These results are consistent with the previous findings using cell lines (16, 17). Similar results were obtained with heterozygous cells. However, the activation by Notch1IC was severalfold lower in homozygous cells in comparison with wild-type and heterozygous cells (Fig. 2C). By the coexpression of Mam-1 and Notch1IC, the promoter was activated to an extent comparable with that of wild-type and heterozygous cells. Similar results were obtained with Notch2IC (data not shown). These results indicate that Mam-1 is therefore necessary to evoke the full activity of Notch signaling in the fibroblasts and that a weak but certain degree of the signaling can thus be transmitted even in the absence of Mam-1. One problem that remains to be solved is whether the residual activity depends on the Mam-2 and/or Mam-3 expressed in the fibroblasts or whether the intracellular domains of Notch can activate their targets in the absence of Mam.

Mam-1^−/− Mice Are Growth Retarded.

The mice deficient in Mam-1 exhibited growth retardation (Fig. 3A). This phenotype is apparent at the perinatal stage, and thereafter it becomes exacerbated in the postnatal period. Around 1 week after birth, Mam-1-deficient mice are 3- to 4-fold smaller than their wild-type or heterozygous littermates (Fig. 3B), and such mice generally succumb before weaning (Fig. 3C). Furthermore, the mice exhibit muscular dystrophy (M. Higashi and K. Azuma, unpublished data). However, the severity of growth retardation, survival, and muscular dystrophy (K. Katsube, unpublished data) are more modest than those of the recently described strain (22), presumably because of difference in the ES cell strains used. The causes of the growth retardation and death are thus currently under investigation.

Fig. 3.

Growth retardation and early death of Mam-1^−/− mice. (A) A Mam-1^−/− mouse and its littermate at P12. (B) Growth of Mam-1^−/− mice and the littermates. (C) Survival of Mam-1^−/− mice. A Kaplan–Meier representation of 10 Mam-1^−/− mice.

Mam-1 Is Dispensable for Generation and Maintenance of Definitive Hematopoiesis.

We examined hematopoietic stem and progenitor cells in the fetal liver at E14.5. We observed no significant difference in the number of both c-Kit⁺Sca-1⁺lineage markers (Lin)⁻ (KSL) hematopoietic stem/progenitor cells and colony-forming cells between the control and Mam-1^−/− fetal liver cells [supporting information (SI) Fig. 6 A–C]. We also examined KSL hematopoietic stem cells and colony-forming cells in bone marrow of mice at postnatal day 7 (P7) to P8. We again observed comparable numbers of these cells in the control and Mam-1^−/− mice when the results were normalized with their body weights (SI Fig. 6 D–F). Finally, we performed a competitive repopulation assay for fetal liver cells from either Mam-1^−/− or control mice with wild-type adult bone marrow cells as competitors at a ratio of 2:1. Nine weeks after this transplantation, the Mam-1^−/− cells were repopulated to a comparable degree with that of the control cells as evaluated based on the percentages of the fetal liver-derived cells among the peripheral blood leukocytes (SI Fig. 6G). These results indicate that Mam-1 is dispensable for development and maintenance of stem cells for definitive hematopoiesis.

Impaired Development of Thymocytes in Mam-1^−/− Mice.

Early T cell development is a recurrent site where mutant mice of the Notch pathway components exhibit phenotypes. At E17.5–E18.5, the number of thymocytes of Mam-1^−/− fetuses was ≈4-fold smaller than those of the control littermates (Fig. 4A). At P7–P8, the number of thymocytes in the Mam-1^−/− mice was smaller by ≈12-fold on average in comparison with those of the control littermates (Fig. 4B). To examine development of thymocytes, we performed a flow cytometric analysis of the CD4 and CD8 phenotypes. As shown in Fig. 4 C and D, the Mam-1^−/− mice in both the developmental stages showed a marked increase (3- to 4-fold at E17.5–E18.5 and 2- to 3-fold at P7–P8) in the proportion of DN fraction in comparison with the control mice.

Fig. 4.

Defective development of early thymocytes in Mam-1^−/− mice. (A) The number of thymocytes at E17.5–E18.5. The data are the mean ± SD (n = 5≈18). (B) The number of thymocytes at P7–P8. The data are the mean ± SD (n = 4≈5). (C) A FACS analysis of thymocytes at E17.5. (D) A FACS analysis of thymocytes at P7. (E–G) A FACS analysis of the thymocytes at P7. The profiles were gated on CD4⁻CD8⁻ cells. (H) A FACS analysis of thymocytes from mixed chimeras. The profiles were gated on CD45.2⁺ cells. The numbers in C–H represent the percentages of cells in the indicated areas. (I) The percentages of DN thymocytes in mixed chimeras. The data are the mean ± SD (n = 3). (J) RT-PCR analysis of mRNAs isolated from fractions of thymocytes. DN, CD4⁻CD8⁻ cells; DP, CD4⁺CD8⁺ cells; CD4SP, CD4⁺CD8⁻ cells; CD8SP, CD4⁻CD8⁺ cells.

We further analyzed the DN fraction of thymocytes at P7 by examining several lineage markers. As shown in Fig. 4E, most of the DN cells from both Mam-1^−/− and the control thymi expressed a T cell marker (Thy1.2) and only a small fraction of the cells express a natural killer cell marker (DX5) to a similar extent. We also performed staging of the DN cells by expression of CD25 and CD44, namely, DN1 (CD44⁺CD25⁻), DN2 (CD44⁺CD25⁺), DN3 (CD44⁻CD25⁺), and DN4 (CD44⁻CD25⁻). As shown in Fig. 4F, the distribution of the DN cells to each of the stages was similar in both the Mam-1^−/− and the control mice. These results indicate that most of the DN cells have characteristics of T cells. Consistently, cells that express B cell markers were not accumulated in thymi of Mam-1^−/− mice (data not shown). Fig. 4G shows that the proportions of the cells that express γδ T cell receptor were similar between Mam-1^−/− and the control thymus. We obtained similar results in the analysis of the DN fraction of thymocytes at E18.5 (data not shown).

We also examined the thymocytes of the mixed bone marrow chimeras. As shown in Fig. 4H, the proportion of the DN fraction was ≈2-fold larger in the Mam-1^−/− fetal liver-derived cells in comparison with the wild-type fetal liver-derived cells. This value is statistically significant (t test, P < 0.05; n = 3) (Fig. 4I). Other phenotypes of the Mam-1^−/− cells colonized in the thymi are reminiscent of those in the young Mam-1^−/− mice (data not shown). These results indicate that Mam-1 is cell-autonomously required for the efficient development of DN thymocytes as Notch1 and RBP-J.

The expression of DNMAML1 in hematopoietic stem cells has been shown to cause a deficiency in the in vivo development of early T cell progenitors (ETPs; CD44⁺CD25⁻c-Kit^hiLin^lo), the most primitive T lineage progenitors in the thymus (27). When we examined the thymi of Mam-1^−/− mice, however, we found that ETPs are present in a number comparable with that in the control thymi (SI Fig. 7). These results indicate that Mam-1 is dispensable for the generation of ETPs.

We analyzed the expression of mRNAs for three Mam genes in the four fractions of thymocytes divided by the expression of CD4 and CD8. As shown in Fig. 4J, not only Mam-1 but also Mam-2 and Mam-3 mRNA were detectable in all four of the fractions by RT-PCR.

Because the total number of thymocytes at P7–P8 is 12-fold smaller and the percentage of DN cells is 2- to 3-fold larger in Mam-1^−/− mice in comparison with the control, the absolute number of DN cells in Mam-1^−/− mice is thus still 4- to 6-fold smaller than the control mice. Because the Mam-1-deficiency does not seem to affect the generation of ETP subset and the development of Mam-1^−/− thymocytes does not seem to become arrested at any of the single DN1–DN4 stages, the rate of development through the whole DN stages is suggested to be reduced in the Mam-1^−/− mice. This idea is in keeping with a result that all of the DN1–DN3 cells require a Notch ligand to develop into DP cells in vitro (13). The steps between the DN stages have also been shown to require the Notch1-dependent signal by an analysis of the competitive repopulation of Notch1^+/+ and Notch1^+/− hematopoietic stem cells (28). Both of these studies also illustrated that stronger Notch1 signal is required to efficiently promote the progression through the DN stages than to suppress B cell development in the thymus (13, 28). Therefore, a weak signal emanating in the absence of Mam-1 may still be sufficient for proper ETP generation and T/B lineage determination but insufficient for the efficient development through the DN stages.

At P7–P8, because the percentage values of DP and single-positive (SP) thymocytes are similar between the Mam-1^−/− mice and the control, there would thus be a ≈12-fold decrease in the absolute number of these cells in Mam-1^−/− mice in comparison with the control. Therefore, at the DN/DP transition, the rate would thus decrease by 2- to 3-fold in the Mam-1^−/− thymocytes. It has recently been shown that the timely expression of DNMAML1 in DN3 stage causes a complete block of DN/DP transition (26). Based on such evidence, it has been proposed that the heterogeneity in the various timings of Cre expression accounts for a partial block in the DN/DP transition observed in the conditional inactivation of Notch1 or RBP-J (26). Therefore, the incomplete block at DN/DP transition observed in Mam-1^−/− thymocytes may be due to a partial failure of the signal generated by Notch1 and RBP-J. In addition, in vitro evidence has also consistently shown that the Notch signal required for this DN/DP transition step is stronger than that needed to suppress the B lineage potential of the progenitor cells (13).

At E17.5–E18.5, the total number of Mam-1^−/− thymocytes is 3- to 4-fold smaller and the percentage of Mam-1^−/− DN cells is 4-fold larger compared with the controls. Therefore, the absolute number of DN cells is comparable between Mam-1^−/− and the control and the obstructing step is therefore mainly considered to be the DN/DP transition. Collectively, these results indicate that Mam-1/Notch1IC/RBP-J complex thus plays a role in early T cell development.

Mam-1 Is Required for the Development of MZB Cells.

We finally examined the transplanted mice for the development of splenic MZB cells, which could not be investigated in the Mam-1^−/− mice per se because of their early death. When assessed by two combinations of markers (B220⁺CD21^hiCD23^lo/− and B220⁺CD1d^hiCD9^hi), we found that proportion of MZB cells was greatly reduced in the Mam-1^−/− cells in comparison with either the control fetal liver-derived cells (Fig. 5A) or the competitor-derived cells colonized in the same spleen (data not shown). These results indicate that Mam-1 is cell-autonomously required for the development of MZB cells as Notch2 and RBP-J, and that the ternary complex formed by these three proteins in response to the stimulation elicited by Delta1 is indispensable for the development of the subset.

Fig. 5.

The defective development of MZB cells from Mam-1^−/− hematopoietic stem cells. (A) A FACS analysis of splenic B cells from mixed chimeras. The profiles were gated on CD45.2⁺B220⁺ cells. The numbers represent the percentages of the cells in the indicated area (MZB cells). (B) An RT-PCR analysis of mRNAs isolated from fractions of splenic B cells. NFB, newly formed B cells; FOB, follicular B cells.

Fig. 5B shows an analysis of mRNAs for the three Mam genes expressed in three fractions of splenic B cells (newly formed B cells, follicular B cells, and MZB cells). All of the Mam mRNA were detectable in all of the fractions by RT-PCR.

Discussion

We have described the generation of Mam-1-deficient mice and the results of an initial analysis for their phenotypes. Mam-1 is responsible for maintaining Notch signaling at its full magnitude in cultured EFs. Among the Notch-dependent stages in hematopoiesis, Mam-1 is (i) required for the development of MZB cells, (ii) partially required for the development from DN to DP thymocytes, and (iii) not required for production of ETPs, T/B bifurcation, or the generation of definitive hematopoiesis. These results suggest that Mam is an essential component of the canonical Notch signaling in mammals in vivo, similar to their counterparts in Drosophila and C. elegans.

The incompleteness in phenocopying the gene-deficient mice of other signaling components may be due to the presence of Mam family proteins, which are functionally similar in the assays involving overexpression (17). However, there is a clear difference in the dependency on Mam-1 between the development of MZB cells and early thymocytes, although all three Mam mRNA can be detected in the cells at both the developmental stages. In the case of EFs, strength of the signal heavily depends on Mam-1 (Fig. 2C), whereas all Mam mRNA can also be detected (Fig. 2B). This kind of bias for dependency on Mam family may also be present in the splenic B cells. Alternatively, the difference may be attributed to differential dependence on strength of intracellular Notch signaling of the processes. As suggested by the haploinsufficiency of Notch2 and Delta1 for the development of MZB cells (11, 12), this process strictly depends on the strength of intracellular Notch signaling. In contrast, haploinsufficiency of Notch1 for thymocyte development becomes apparent only in the presence of wild-type cells as competitors (28, 29), implying the dependency on Notch signal of this process is weaker than that of MZB development. These two possibilities are not mutually exclusive. In this sense, it is worth noting that a small number of MZB cells were in fact present in all of the examined splenocytes derived from Mam-1^−/− fetal liver cells (n = 3; data not shown). These cells may have developed with the help from Mam-2 and/or Mam-3. Nevertheless, there still is a possibility that residual activity evoked in the absence of Mam is sufficient to implement some Notch-dependent steps in vivo.

Overexpression of several Notch inhibitors in early lymphoid progenitors resulted in various but partially overlapping outcomes (25, 30–32). Among the inhibitors, the N-terminal basic domain of Mam-1 (DNMAML1) seems to nonspecifically inhibit the Notch-dependent steps and is proposed as a pan-Notch inhibitor. Indeed, the phenotypes produced by DNMAML1 are stronger in comparison with those by the deficiency of Mam-1 (25–27); thus the present study may further endorse this notion. It has been postulated that the C-side portions of Mam serve as a major scaffold for the transcriptional coactivators in the Mam–NotchIC–CSL complex and that a transcriptionally inert complex emerges by deleting this portion. However, this model is contradictory to our results and those of others in which the strength of transcription largely depends on NotchIC but not on that of Mam in transcriptional transactivation assays using combinations of four kinds of mammalian Notch and the three kinds of Mam (17, 19, 33). These results indicate that NotchIC but not Mam mainly behaves as a coactivator for transcription from the promoter (33). Therefore, the C-side portions of Mam may somehow support the activation domains of Notch to work properly. A model that is consistent with the currently available results would be that the association of NotchIC with CSL is destabilized by sequestering endogenous Mam by the unaccompanied N-terminal basic domain that can only poorly stabilize NotchIC–CSL in comparison with its full-length counterpart (16), which is essential for the complex formation (present study).

Methods

Targeting Mam-1 Allele.

A genomic clone containing Mam-1 was isolated by screening a library made with genomic DNA from a mouse of 129 strain with a cDNA encoding murine Mam-1 (IMAGE clone; 458724) as a probe. A targeting vector was then constructed with pMC1NeoPolyA. A diphtheria toxin A gene cassette was also ligated at the 3′ end of the targeting vector. The vector was linearized at the 5′ end and then introduced into E14-1 ES cells by electroporation. After selection in a medium containing G418, clones with a targeted allele were identified by Southern blotting using a standard procedure. The targeting frequency was ≈2%. ES cells with the targeted allele were injected into blastocysts of C57BL/6 strain. Chimeric males were mated with females of the C57BL/6 strain to obtain heterozygous mice. All animals were genotyped by either Southern blotting or by PCR. Germline transmission was obtained from one of the clones. The mice of the F₂ generation were used for the experiments described in this report except for a competitive repopulation assay of the hematopoietic stem cells. All mice were kept under specific pathogen-free conditions. All experiments with animals were approved by the Committee of Animal Experiments, Chiba University Graduate School of Medicine.

Cell Culture and Luciferase Assay.

EFs were prepared from E14.5 embryos and cultured in Dulbecco's modified Eagle's medium supplemented with 10% FBS.

The cells seeded on 12-well plates were then transfected with pEF-BOS with or without Mam-1 (16), pEF-BOSneo with or without Notch1IC (RAMIC) (17, 34), pTP1-luc (35), and Renilla luciferase internal control plasmid (pRL-CMV; Promega, Madison, WI) by using LipofectAMINE (Invitrogen, San Diego, CA). Two days after the transfection, the firefly and Renilla luciferase activities were determined by using a Dual Luciferase Assay kit (Promega) and a Turner Designs (Palo Alto, CA) TD20/20 dual luminometer. Firefly luciferase activities were normalized with the Renilla luciferase control activities.

Western Blot Analysis.

Whole-cell extracts were prepared from EFs as previously described (16). After separation by SDS/PAGE, the proteins were electrophoretically transferred onto polyvinylidene difluoride membranes (Bio-Rad, Hercules, CA). The antibodies used for blotting were anti-Mam-1 (17), anti-Notch1 (Santa Cruz, Santa Cruz, CA; sc-6014), anti-Notch2 (Santa Cruz; sc-5545), and anti-Hes1 [gift from T. Sudo (Toray Pharmaceutical Research Laboratories, Kamakura, Japan)]. They were visualized with appropriate secondary antibodies conjugated with horseradish peroxidase and a Lumi-Light Plus (Roche, Basel, Switzerland) chemiluminescent substrate by using a LAS-1000plus luminescent image analyzer (FUJIFILM, Tokyo, Japan).

Quantitation of mRNA.

For a real-time PCR analysis, total RNA was isolated by using TRIzol reagent (Invitrogen). Reverse transcription of the RNA was done with ExScript RT reagent kit (Takara Bio, Otsu, Japan). Real-time PCR was performed with SYBR Green I by the SYBR premix Ex Taq (Perfect Real-Time) kit (Takara Bio) and the ABI PRISM 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). The expression of rps18 was used as an internal control.

Semiquantitative RT-PCR was carried out by using normalized cDNA by quantitative PCR with TaqMan rodent GAPDH control reagent (Applied Biosystems) as previously described (36). The PCR products were separated on agarose gel and visualized by ethidium bromide staining. DN, DP, CD4SP, and CD8SP cells were isolated as previously described (36). Newly formed B cells, follicular B cells, and MZB cells were sorted as B220⁺CD21^lo/−CD23^lo/−, B220⁺CD21^int-hiCD23^hi, and B220⁺CD21^hiCD23^lo/− splenocytes, respectively, from C57BL/6 mice with JSAN cell sorter (Bay bioscience, Kobe, Japan).

The primer sequences and amplification conditions are available from the authors on request.

Flow Cytometric Analysis.

Reagents used for analyzing the development of thymocytes and splenocytes were as follows. Phycoerythrin (PE)-CD4, allophycocyanin (APC)-CD8, PE-DX5, biotin-CD4, FITC-conjugated TCRαβ, PE-TCRγδ, biotin-CD45.2, FITC-CD21, biotin-CD9, FITC-CD45.2, and PerCp-Cy5.5-CD3 antibodies and streptavidin-conjugated PerCp-Cy5.5 were purchased from BD Biosciences (Mountain View, CA). FITC-Thy-1.2, APC-CD8, FITC-CD44, PE-CD25, PE-CD23, APC-B220, PE-CD1d, APC-c-Kit, biotin-Gr-1, biotin-Mac-1, biotin-B220, and biotin-CD8 were purchased from eBioscience (San Diego, CA). After staining, the cells were analyzed with FACSCalibur (BD Biosciences).

Abbreviations

Mam

Mastermind

IC

intracellular

CSL

CBF-1

Su(H)

Lag-1

RBP-J

recombination signal sequence-binding protein J

DP

double positive

DN

double negative

SP

single positive

MZB

marginal zone B

EF

embryonic fibroblast

E14.5

embryonic day 14.5

P7

postnatal day 7

ETP

early T cell progenitor

PE

phycoerythrin.