The muscarinic gene family was the first complete subfamily of GPCRs to be cloned (Bonner et al. 1987, 1988; and see chapter 18). There are five members, all of which have unique but partially overlapping expression profiles (Buckley et al. 1988; Vilaro et al. 1990). The M4 gene has a simple gene structure consisting of a single non-coding exon (containing two transcription initiation sites) and a single coding exon (Mieda et al. 1996, 1997; Wood et al. 1996, 1995). Sequence analysis of the 5'flanking region of this gene revealed the presence of a RE1/neuron restrictive silence element (NRSE) (Kraner et al. 1992; Mori et al. 1992) situated 0.5 kb upstream of the first transcription initiation site (Mieda et al. 1997; Wood et al. 1996) The RE1/NRSE is a 21 bp motif that binds a Kruppel type zinc finger transcriptional repressor REST/NRSF. Both the RE1/NRSE and REST/NRSF were originally characterized by the groups of Gail Mandel at Stony Brook and David Anderson at Caltech
(Chong etal. 1995; Schoenherr and Anderson 1995a). Based on transient transfection assays and expression studies the consensus evolved that REST/NRSF acted to silence expression of NRSE-bearing neuron-specific genes such as the SCG10 and Na type II sodium channel in non-neuronal tissue and in undifferentiated neuroepithelial cells (Schoenherr and Anderson 1995b; Schoenherr et al. 1996). The role of the NRSEM4 was tested within this context. As predicted, removal of the NRSE from reporter gene constructs led to activation of transcription from the M4 promoter only in NRSF-expressing cells. Likewise overexpression of NRSF led to repression of NRSE-bearing constructs while expression of dominant-negative constructs containing only the NRSF DNA binding domain, alleviated the repression. Elec-tromobility shift assays (EMSA) also confirmed the ability of NRSF to interact with the NRSE of the M4 gene (Mieda et al. 1997; Wood et al. 1996). These data were interpreted as showing that NRSF acts to silence expression of the M4 gene in NRSF-expressing cells and were taken to re-enforce the idea that NRSF acts as a silencer of NRSE-bearing genes outside of the differentiated nervous system. However, all of these studies address the singular issue of potential—can NRSF interact with the NRSEM4 to repress transcription from the M4 promoter? None of these studies examined the interaction of NRSF with the endogenous NRSEM4—does NRSF interact with the endogenous M4 gene? When this is done a more complicated picture emerges.
We have examined the interaction of NRSF with active and silent endogenous M4 loci. We have used rat JTC-19 fibroblasts which express endogenous NRSF and do not express M4 receptors, and PC12 pheochromocytoma cells which express M4 receptors and are one of the few cell lines that do not express NRSF (or at least very little) and into which we introduced recombinant NRSF constructs. DNase hypersensitivity, endonuclease access assays and chromatin immunoprecipitation all failed to show any evidence of interaction between NRSF and the NRSEM4at the silent M4 locus in JTC-19. In contrast, these same assays showed that recombinant NRSF repressed expression of the endogenous M4 gene and did interact with the chromatin around the NRSEM4 of the active locus in PC12 cells. These data are more consistent with NRSF acting as a modulator of expression levels rather than being required for maintenance of the silenced state—quite different from the conclusions drawn from transient transfections of reporter gene constructs. Consistent with this view is the observation that relief of histone deacetylase activity by Trichostatin A (TSA) does not induce expression of the M4 gene in JTC-19 cells (Wood and Buckley, unpublished observations). However, this picture is further complicated when another NRSE-bearing gene, the Na type II channel gene is examined. In this case, chromatin immunoprecipitation establishes that NRSF is associated with the NRSENan and accordingly, expression of the Na type II gene is induced by TSA (Ballas et al. 2001). It is important to note that transient reporter gene analysis of these two genes indicated that both the M4 and Na type II genes are NRSF responsive. Several other GPCR genes have NRSEs in their regulatory region including a rat CCR10 receptor (Bonini et al. 1997), two orphan receptors; the human GPR10 (Marchese etal. 1995) and the GPR6 receptor (accession U18549) and the mouse ^ opioid (Pan et al. 1999). As yet, their functional significance has not been assessed. The lesson seems clear: the existence of a particular regulatory element in a GPCR gene does not necessarily indicate that it acts to recruit its cognate transcription factor. Only detailed analysis of the chromatin environment of individual GPCR genes will solve this issue. It is of course possible that NRSF might be required to establish but not to maintain the silenced state—this has a parallel in Drosophila where two Polycomb group complexes exist—one to establish and one to maintain silence (Farkas etal. 2000).
Another line of evidence pointing to a more refined role for NRSF than initially assumed comes from experiments on NRSF-/- mice. Although no GPCR was examined in these studies, ablation of the NRSF gene did not lead to widespread expression of NRSE-bearing genes in non-neural tissues. Only piII-tubulin expression showed prolonged up-regulation in areas where it is normally transiently expressed (Chen et al. 1998). One possibility that arises from these observations is that some NRSE-bearing GPCR genes may require NRSF to establish or maintain silence while others may not.
How does this NRSF-mediated regulation of the M4 gene occur? Several studies have shown that the N-terminal repression domain of NRSF interacts with the sin3 co-repressor while the C-terminal recruits a distinct co-repressor that contains Co-REST but not sin3 (Andres et al. 1999; Grimes et al. 2000; Huang et al. 1999; Roopra et al. 2000). Both co-repressor complexes interact with histone deacteylases which can remove the N-terminal acetyl groups from histone tails of H3 and H4 causing condensation of the nucleosome and consequent repression of transcription. This remodelling is accompanied by an increase in DNase hypersensitivity, endonuclease access and formation of ordered nucleosomes around the NRSE (Wood and Buckley unpublished observations). There is no reason to think that these experiments are anything but typical in highlighting the apparent disparate roles attributed to particular transcription factors in regulating GPCR expression when results from transient expression of reporter genes in cultured cells are compared with those carried out on endogenous genes.
Was this article helpful?