The transcriptional repressor REST determines the cell-specific expression of the human MAPK8IP1 gene encoding IB1 (JIP-1)

Abstract

Islet-brain 1 (IB1) is the human and rat homologue of JIP-1, a scaffold protein interacting with the c-Jun amino-terminal kinase (JNK). IB1 expression is mostly restricted to the endocrine pancreas and to the central nervous system. Herein, we explored the transcriptional mechanism responsible for this preferential islet and neuronal expression of IB1. A 731-bp fragment of the 5' regulatory region of the human MAPK8IP1 gene was isolated from a human BAC library and cloned upstream of a luciferase reporter gene. This construct drove high transcriptional activity in both insulin-secreting and neuron-like cells but not in unrelated cell lines. Sequence analysis of this promoter region revealed the presence of a neuron-restrictive silencer element (NRSE) known to bind repressor zinc finger protein REST. This factor is not expressed in insulin-secreting and neuron-like cells. By mobility shift assay, we confirmed that REST binds to the NRSE present in the IB1 promoter. Once transiently transfected in beta-cell lines, the expression vector encoding REST repressed IB1 transcriptional activity. The introduction of a mutated NRSE in the 5' regulating region of the IB1 gene abolished the repression activity driven by REST in insulin-secreting beta cells and relieved the low transcriptional activity of IB1 observed in unrelated cells. Moreover, transfection in non-beta and nonneuronal cell lines of an expression vector encoding REST lacking its transcriptional repression domain relieved IB1 promoter activity. Last, the REST-mediated repression of IB1 could be abolished by trichostatin A, indicating that deacetylase activity is required to allow REST repression. Taken together, these data establish a critical role for REST in the control of the tissue-specific expression of the human IB1 gene.

FIG. 1

Nucleic acid sequence of a fragment of the human IB1 promoter. The transcription start site (asterisk, +1) was determined by a 5′RACE reaction and computer analysis. A putative NRSE binding site is underlined. The ATG in boldface is the codon of translation initiation of the IB1 protein.

FIG. 2

Expression of the IB1 and REST genes in several cell lines. (A) Northern blot analysis of IB1 and β-actin mRNAs in insulin-secreting cells (βTC3 and INS1), in rat neuron-like PC12 cells, in human carcinoma HeLa cells, in mouse fibroblast-derived NIH 3T3 cells, in human lymphoma Jurkat cells, and in mouse macrophage-like RAW cells. Total RNA (15 μg) was hybridized with the human IB1 cDNA probe as described previously (4). β-Actin was used to control the quantity of the total RNA loaded. (B) Analysis of REST expression in βTC3, INS1, PC12, HeLa, NIH 3T3, Jurkat, and RAW cells. RT-PCR analysis of RNA from different cell lines yielded one detectable PCR fragment using specific primer pairs for human and mouse and rat REST mRNA (primer pairs comprising primers 1 and 2 and 3 and 4, respectively). These primers, designed as described previously (26), were between exon 4 and exon 6 to span the region which differs in REST mRNA isoforms. Sequence analysis of the amplified fragment showed that the PCR product of HeLa, NIH 3T3, Jurkat, and RAW cells is the REST full-length cDNA fragment. This band was not detectable in βTC3, INS1, and PC12 cells. (Bottom) Ubiquitous cytochrome oxidase 4 (COX4) was also amplified to control the quality of the reverse transcription for each mRNA sample. PCR was performed as described in Materials and Methods. −, negative control for which no reverse transcriptase was added to the reaction.

FIG. 3

Transcriptional activity of the IB1 promoter region. The region of the IB1 promoter comprising bp −691 to + 41 (IB1LUC) drove a high luciferase activity in the insulin-secreting βTC3, INS1, and PC12 cells (13-, 11-, and 18-fold higher, respectively, than that for the promoterless vector pGL3 basic) but not in unrelated HeLa, NIH 3T3, Jurkat, and RAW cells. Each experiment was repeated at least three times in triplicate. Luciferase activities were normalized using pRLCMVrenilla, and results are expressed as means ± standard errors of the means (triple asterisk, P < 0.001).

FIG. 4

Sequence-specific binding activity of REST to the human IB1 NRSE. (A) EMSA with ³²P-labeled IB1 NRSE using nuclear extracts from HeLa cells. A slow-migrating complex (arrowhead A) was detected, compared to free-probe migration (arrowhead B). This pattern was supershifted by using REST antibodies (arrowhead C). (B) The DNA-binding activity with IB1 NRSE was competed by adding a 100- or 600-fold molar excess of unlabeled wild-type NRSE (Wt) but not by adding unlabeled mutated NRSE (Mut). NS, nonspecific competitor.

FIG. 5

Mutation of the NRSE motif in the human IB1 promoter relieved transcriptional activity in REST-expressing cell lines. The IB1 promoter containing mutated or wild-type NRSE, mIB1LUC and IB1LUC, respectively, was transiently transfected into REST-expressing HeLa, NIH 3T3, Jurkat, and RAW cells. The luciferase activity was relieved in all these cells with mIB1LUC. Each experiment was performed at least three times in triplicate. Luciferase activities were normalized using pRLCMVrenilla, and results are expressed as means ± standard errors of the means (triple asterisk, P < 0.001). RLU, relative light units.

FIG. 6

REST represses IB1 promoter activity. IB1LUC and mIB1LUC constructs were cotransfected in the insulin-secreting cell line (βTC3), with REST expression vector RIP REST (under the control of the RIP) or the empty RIP vector as a control. The luciferase activity of IB1LUC was decreased by 50% in REST-expressing cells, whereas REST expression had no effect on mIB1LUC activity. Each experiment was performed at least three times in triplicate. Luciferase activities were normalized using pRLCMVrenilla, and results are expressed as means ± standard errors of the means (asterisk, P < 0.05). RLU, relative light units.

FIG. 7

Human NRSE silences the activity of the heterologous SV40 promoter. The wild-type and mutated NRSEs of the IB1 promoter, WTNRSE and MUTNRSE, respectively, were cloned as dimers upstream an SV40 promoter. Constructs were then transiently transfected into HeLa, NIH 3T3, Jurkat, and RAW cells. WTNRSE activity is markedly decreased in all these cells compared to the basal activity of SV40LUC. The mutation in NRSE partially restored the activity of the MUTNRSE promoter. Luciferase activities were normalized using pRLCMVrenilla. Each experiment was performed at least three times in triplicate. All values are expressed as percentages of the SV40LUC activity and are means ± standard errors of the means (triple asterisk, P < 0.001).

FIG. 8

Expression of DNREST derepressed the IB1 promoter and heterologous SV40 promoter activities in HeLa cells. (A) IB1LUC and mIB1LUC vectors were transiently transfected with or without DNREST into HeLa, NIH 3T3, Jurkat, and RAW cells. The IB1LUC activity was relieved with the expression of DNREST. Triple asterisk, P < 0.001. RLU, relative light units. (B) The heterologous construction WTNRSE was transfected in HeLa, NIH 3T3, Jurkat, and RAW cells with or without DNREST. The WTNRSE activity was dramatically decreased compared to the SV40LUC basal activity. The WTNRSE activity was completely restored upon DNREST expression. Triple asterisk, P < 0.001.

FIG. 9

IB1 promoter activity is repressed through NRSE in a TSA-sensitive manner. (A) IB1 promoter construct IB1LUC was transiently transfected into HeLa, NIH 3T3, Jurkat, and RAW cells. TSA (100 nM) or DMSO as a control was added to medium 24 h after transfection. The IB1LUC activity was relieved in all TSA-treated cells. (B) The mIB1LUC construct was also transfected in the REST-expressing cells with TSA or DMSO. No significant differences in mIB1Luc activity between TSA-treated cells and DMSO-treated cells were observed. (C) The heterologous WTNRSE construct was also transiently transfected in the presence of TSA or DMSO. The WTNRSE activity was restored in all TSA-treated cells. Each experiment was performed at least three times in triplicate. Triple asterisk, P < 0.001. RLU, relative light units.

FIG. 10

Endogenous IB1 gene expression is relieved in TSA-treated cells. Total RNAs from HeLa, NIH 3T3, Jurkat, and RAW cells treated in the presence of DMSO (−) or 100 nM TSA (+) were analyzed for IB1 gene expression by RT-PCR using specific primer pairs as described in Materials and Methods. RNA from βTC3 cells was used as a positive control for IB1 gene expression. In REST-expressing cells, the PCR yielded one PCR fragment detected only in the presence of TSA-treated cells.