Repression of c-myc Transcription by Blimp-1, an Inducer of Terminal B Cell Differentiation

See allHide authors and affiliations

Science  25 Apr 1997:
Vol. 276, Issue 5312, pp. 596-599
DOI: 10.1126/science.276.5312.596


Transcription of c-myc in plasma cells, which are terminally differentiated B cells, is repressed by plasmacytoma repressor factor. This factor was identified as Blimp-1, known for its ability to induce B cell differentiation. Blimp-1 repressed c-myc promoter activity in a binding site–dependent manner. Treatment of BCL1 lymphoma cells with interleukin-2 (IL-2) plus IL-5 induced Blimp-1 and caused a subsequent decline in c-Myc protein. Ectopic expression of Blimp-1 in Abelson-transformed precursor B cells repressed endogenous c-Myc and caused apoptosis; Blimp-1–induced death was partially overcome by ectopic expression of c-Myc. Thus, repression of c-myc is a component of the Blimp-1 program of terminal B cell differentiation.

c-Myc functions at a critical decision point of cell growth to favor proliferation and to block terminal differentiation (1). c-Myc is present in dividing cells but is not expressed in quiescent or terminally differentiated cells (2); addition of exogenous c-Myc blocks terminal differentiation of several hematopoietic cell lines (3) and of myogenic cells (4), whereas inhibitors of c-Myc expression accelerate terminal differentiation of promonocytic HL60 cells (5), M1 leukemic myeloid cells (6), F9 teratocarcinoma cells (7), and human esophageal cancer cells (8).

Murine plasmacytomas are the transformed counterparts of plasma cells, which are terminally differentiated, nondividing B lymphocytes (9). Plasmacytomas have a characteristic reciprocal chromosomal translocation that juxtaposes one allele of the c-myc gene with an immunoglobulin heavy- or light-chain locus (10). The translocated c-myc allele is deregulated and overexpressed; however, the nontranslocated c-myc allele is transcriptionally silent (1). This state is thought to correspond to the silent state of the c-myc gene in normal plasma cells.

A plasmacytoma-specific protein, plasmacytoma repressor factor (PRF), binds in the c-myc promoter 290 base pairs (bp) 5′ of the P1 transcriptional start site. PRF represses c-myctranscription in plasmacytomas and has not yet been cloned (11). The PRF binding site is similar in sequence to the interferon-stimulated response elements (ISREs) in many interferon-regulated genes (12) and to the positive regulatory domain 1 (PRD1) sequence of the human interferon-β (IFN-β) gene (13). Electrophoretic mobility shifts assays (EMSAs) with nuclear extracts from the plasmacytoma P3X63-Ag8 (P3X) and a c-myc promoter probe containing the PRF site confirmed that both ISRE and PRD1 oligonucleotides could compete for binding of PRF in this assay; PRD1 oligonucleotides competed more strongly than ISRE oligonucleotides (14).

PRD1-BF1 is a human zinc finger protein that was cloned by virtue of its ability to bind to the PRD1 site; PRD1-BF1 inhibits transcription of the IFN-β promoter (13). Recently the murine homolog of PRD1-BF1, Blimp-1, was identified as a protein that is induced upon stimulation of the BCL1 B cell lymphoma line with interleukin-2 (IL-2) plus IL-5 (15). Ectopic expression of Blimp-1 can drive B cell terminal differentiation, and Blimp-1 is expressed only in plasmacytomas and mature B cells; however, its mechanism of action is not well understood (15). On the basis of cross-competition of their binding sites, common transcriptional repressor activity, and plasmacytoma-specific expression, we hypothesized that PRF might be identical to Blimp-1.

To test this hypothesis, we transfected 293T human kidney fibroblast cells with an expression plasmid encoding Blimp-1. An immunoblot developed with antiserum to Blimp-1 revealed that Blimp-1 was present in nuclear extracts from P3X plasmacytomas and in the transfected 293T cells but not in 18-81 precursor B cells (pre-B cells) or in mock-transfected 293T cells (Fig. 1). EMSAs were then done with these extracts with an oligonucleotide probe corresponding to the c-myc PRF site (Fig. 1). Complexes of identical mobility were observed for endogenous PRF in P3X cells and for the Blimp-1–transfected 293T cells, whereas no complex was detected for 18-81 or mock-transfected 293T cell extracts. The sequence specificity of these complexes was shown by the ability of PRF but not an unrelated sequence to compete the complexes. Finally, the complex from P3X extracts was completely ablated by antiserum against Blimp-1 but not by naı̈ve antiserum. Thus, the protein in P3X cells that we identified as PRF is immunologically related to Blimp-1. The results obtained with EMSA and antibody ablation provide evidence that the c-myc repressor PRF is encoded by the blimp-1gene.

Figure 1

Blimp-1 binds to the c-myc PRF site. (Top) Nuclear extracts from various cells were prepared as described (26) and subjected to immunoblot with antibody to Blimp-1; arrow indicates Blimp-1. Lane 1, 18-81 cells; lane 2, P3X cells; lane 3, mock-transfected 293T cells; lane 4, Blimp-1–transfected 293T cells. (Bottom) Lanes 5 to 8, the same extracts were used for EMSA with a 25-bp PRF oligonucleotide (26). Lanes 9 to 13, EMSA of P3X nuclear extracts with PRF oligonucleotide probe in the presence of no competitor (lane 9), PRF oligo tetramer (lane 10), GATA (nonspecific) tetramer (lane 11), rabbit antiserum to Blimp-1 (lane 12), and naı̈ve rabbit serum (lane 13). Arrow indicates the protein-DNA complexes.

A site-directed mutation in the c-myc PRF site increases promoter activity 30-fold in plasmacytomas, which express PRF, but has no effect in fibroblasts and pre-B cells, which do not express the protein, showing that PRF represses c-myc transcription (11). To investigate the functional relation between PRF and Blimp-1, we tested the effect of ectopically expressed Blimp-1 on the activity of the c-myc promoter. Reporters dependent on a c-myc promoter with either wild-type or mutated PRF sites were cotransfected with a Blimp-1 expression plasmid into the 18-81 and 300-18 pre-B cell lines (Fig. 2, A and B, respectively). In both pre-B cell lines, wild-type and mutant promoters had similar activity in the absence of Blimp-1. Expression of Blimp-1 repressed the wild-type promoter by 70% in both cell lines, whereas the promoter harboring a mutation in the PRF site was not repressed (Fig. 2, A and B). Thus, ectopic Blimp-1 represses c-myctranscription in pre-B cells, and the repression depends on the presence of the PRF site at −290 bp. Because the function of Blimp-1 is the same as that previously shown for PRF, these data, in conjunction with the in vitro binding studies, suggest that PRF is encoded by the blimp-1 gene.

Figure 2

Blimp-1 represses the c-myc promoter in a PRF site–dependent manner in pre-B cells. Two micrograms of a luciferase reporter fused with either a wild-type (myc-Luc) or PRF-deleted (m-myc-Luc) c-myc promoter including −1100 to +580 bp from P1 (11) with 10 μg of Blimp-1 expression vector (pBDP1-F) or expression vector control (pBDP1-B) (15) were used to transfect 18-81 or 300-18 pre-B cells. Log-phase cells (5 × 106) were collected by centrifugation and resuspended in 300 μl of culture medium, mixed with DNA, and subjected to electroporation at 960 μF and 240 V. Luciferase activity was measured 18 hours after transfection. Results show the average of at least three independent transfections. Error bars show standard deviations. (A) 18-81 cells. (B) 300-18 cells.

Blimp-1 is a nuclear protein with five zinc fingers in its COOH-terminus (15). Only two other target genes for Blimp-1 are known. Human Blimp-1 (PRD1-BF1) represses IFN-β transcription (13) and Blimp-1 activates J-chain transcription, although a binding site in the J-chain gene has not been identified (15). Transcriptional activation and repression in different gene contexts is frequently associated with zinc finger proteins including YY1 (16), Krüppel (17), and glucocorticoid receptor (18) and is often determined by binding of adjacent proteins. The Blimp-1 binding site in the c-myc promoter is near a YY1 binding site that functions as an activator site (19). Our preliminary data show that Blimp-1 and YY1 associate in vitro (20), and it may be that Blimp-1 and YY1 interfere with one another’s activity in the c-myc promoter. However, further studies will be required to understand the mechanisms that determine the activity of Blimp-1 on this and other promoters.

BCL1 is a mature B cell line that, upon stimulation with IL-2 plus IL-5, differentiates into a plasma cell–like state (21). Blimp-1 is induced early in BCL1differentiation (15). On the basis of the ability of transfected Blimp-1 to repress the c-myc promoter, we predicted that differentiation dependent on induction of Blimp-1 in BCL1 cells would cause a decrease in endogenous c-Myc. Seventy-two hours after BCL1 cells were treated with IL-2 plus IL-5, differentiation was verified by increased immunoglobulin secretion and changes in cell size as indicated by changes in forward versus orthogonal scatter (14). c-Myc amounts during this period were assessed by immunoblot (Fig. 3A). After a transient increase, c-Myc decreased by ∼75% between 1 and 2 hours of IL-2 plus IL-5 stimulation and remained low for 72 hours. Northern (RNA) analyses showed that 1 hour after stimulation, Blimp-1mRNA increased approximately five times (15) (Fig. 3B). These data are consistent with the notion that Blimp-1 represses endogenous c-myc transcription, resulting in decreased c-Myc protein.

Figure 3

Kinetics of Blimp-1 induction and c-Myc reduction during BCL1 cell differentiation. (A) BCL1 cells were treated with IL-2 + IL-5; whole-cell extracts were prepared at various times (15), and protein concentrations were measured by the Bradford assay (27). Ten micrograms of protein were subjected to electrophoresis on SDS-polyacrylamide gels (8%), transferred to a nitrocellulose membrane, and immunoblotted with polyclonal antiserum raised against the COOH-terminus of murine c-Myc. The bands were quantitated, and the amounts relative to that at 0 hours are shown. Several repeated experiments gave similar results. (B) RNA was also prepared and analyzed by Northern blotting with ablimp-1 cDNA probe (15) and a control β-actin probe. The relative amounts of blimp-1 mRNA are given below the lanes.

Because Blimp-1 drives terminal B cell differentiation, we reasoned that the c-myc gene would be an important target for Blimp-1–mediated repression. To test this hypothesis directly, we transfected 18-81 pre-B cells with combinations of a neo–Blimp-1 expression plasmid, or a Blimp-1 antisense control, and the pSV2 c-myc expression plasmid, or a pSV2 control. Few colonies were obtained with the Blimp-1 expression plasmid (Fig.4), and 16 of 16 analyzed did not express Blimp-1 (14). However, transfections with varying ratios of Blimp-1 and c-myc plasmids gave 17 to 65% the number of control colonies; four of five colonies analyzed expressed Blimp-1 (14). Thus, ectopic c-Myc blocked the growth-suppressing effect of high Blimp-1 expression, suggesting that repression of c-myc transcription by overexpressed Blimp-1 is directly responsible for the death of Blimp-1–overexpressing cells.

Figure 4

c-Myc blocks the growth suppression effect of high Blimp-1 expression. A pBJ-neo plasmid containing either antisense or sense blimp-1 cDNA was cotransfected into 18-81 pre-B cells with the pSV2-myc expression plasmid or a pSV2 control. Cells were diluted into 96-well plates, cultured with G418 (800 μg/ml), and resistant colonies were counted 10 days later. Colony numbers obtained for different ratios of Blimp-1 sense (S) or antisense (AS) to c-myc plasmids were as follows: (i) 1:2 control (5 μg of Blimp-1:10 μg of pSV2): S = 34 ± 7, AS = 1033 ± 46; (ii) 1:2 (5 μg of Blimp-1:10 μg of pSV2-myc): S = 347 ± 73, AS = 2011 ± 153; (iii) 1:5 (2 μg of Blimp-1:10 μg of pSV2-myc): S = 340 ± 11, AS = 599 ± 24; and (iv) 1:10 (2 μg of Blimp-1:20 μg of pSV2-myc): S = 446 ± 30, AS = 684 ± 19. A graphic representation was generated from these data; for each transfection with the Blimp-1 plasmid, the corresponding transfection with antisense Blimp-1 was taken as 100%.

To obtain direct evidence that ectopic Blimp-1 can repress the endogenous c-myc gene, we made stable B cell transfectants in which Blimp-1 was controlled by a metallothionein promoter. Cadmium treatment of 18-81 transfectants induced Blimp-1 mRNA more than 10 times relative to controls and reduced endogenous c-Myc by more than 90% with similar kinetics (Fig. 5A). This result shows that Blimp-1 directly represses c-Myc expression in vivo. In addition, 18-81 cells expressing Blimp-1 died by apoptosis, as shown by the decrease in viable cells (Fig. 5B) and by fragmentation of nuclear DNA revealed by gel electrophoresis (14) and terminal deoxynucleotidyl transferase–mediated dUTP nick-end labeling (TUNEL) assay (Fig. 5C). However, induction of Blimp-1 in BCL1transfectants revealed terminal differentiation including increased surface expression of Syndecan (14) and secretion of immunoglobulin M (IgM) (Fig. 5D). The reason Blimp-1 caused apoptosis rather than differentiation in 18-81 cells was probably because Abelson murine leukemia virus (AMuLv)–transformed pre-B cells cannot activate nuclear factor kappa B, which is required for immunoglobulin light-chain expression (22). Apoptosis induced by decreased c-Myc has also been reported in WEHI 231 (23) and Ramos (24) B cell lines.

Figure 5

Induction of Blimp-1 alters properties of 18-81 and BCL1 cells. 18-81 and BCL1 cells were stably transfected with an expression vector in which blimp-1 expression was dependent on the sheep metallothionein promoter (MT-B) (28) or an empty control (MT). (A) (Upper panel) Northern blot of Blimp-1 mRNA after treatment of a MT-B 18-81 transfectant with 20 μM cadmium. (Lower panel) Immunoblot from the same cells showing c-Myc after treatment with 20 μM cadmium. (B) Viability of 18-81 MT-B transfectant (•) and MT control (○), determined by trypan blue exclusion, after growth in varying concentrations of cadmium for 36 hours. (C) Detection of Blimp-1–induced apoptosis in 18-81 transfectants by TUNEL staining (29). 18-81 cells transfected either by MT-B or MT control vectors were grown without (−Cd) or with (+Cd) 20 μM cadmium for 36 hours, centrifuged onto slides, and then fixed with 4% paraformaldehyde. The TUNEL staining was done according to the manufacturer’s instructions (Boehringer Mannheim). Arrows indicate apoptotic cells. (D) Cells (105) from BCL1 transfectants containing MT or MT-B (MT-B1 to 3) were treated with 20 μM cadmium for 72 hours, and IgM in the culture medium was measured by enzyme-linked immunosorbent assay (15). Untransfected BCL1 ± IL-2/IL-5 served as controls.

Although c-Myc’s role in blocking terminal differentiation has been well established, we have now shown that Blimp-1 specifically represses c-myc transcription as part of a program of terminal B cell differentiation. Our data show Blimp-1–dependent repression of c-myc and emphasize that growth and differentiation of B cells are exquisitely sensitive to c-Myc. It will be important to determine if Blimp-1 has other important targets or if blocking proliferation by way of c-myc repression is sufficient to trigger terminal B cell differentiation. It will also be interesting to determine if immunoglobulin gene sequences overcome Blimp-1–dependent repression of c-myc transcription in plasmacytomas where the Blimp-1 site is not removed by translocation.

The restricted pattern of Blimp-1 expression (11,15) suggests that the Blimp-1–mediated suppression of c-myc transcription may be unique to B lymphocytes. However, the human homolog PRD1-BF1 is induced upon viral infection of fibroblasts (13), and preliminary results show that Blimp-1–deficient mice die during embryonic development (25). Therefore, it will be important to determine if Blimp-1 also represses c-myc transcription in other lineages of terminally differentiated cells.


View Abstract

Navigate This Article