The proenkephalin gene is a well-studied model of transcription factor-target gene interaction in the nervous system and has been proposed as a regulatory target of the protein product of the immediate-eariy gene c-fos. This regulatory mechanism has been proposed, in part, because the cAMP response element 2 (CRE-2) site, the key DNA regulatory element within the proenkephalin second-messenger-inducible enhancer, avidly binds AP-1 proteins, incuding Fos, in vitro. However, we observe a dissociation in the time course of activation of c-fos and proenkephalin mRNA in rat striatum after adminstration of the dopamine D2 receptor antagonist haloperidol. This result prompted us to investigate the composition of protein complexes in striatal nuclear extracts that bind to the CRE-2 site. Even though our striatal nuclear extracts had substantial basal and haloperidol-inducible AP-1-binding activities that contained Fos, we could not detect Fos in complexes bound to the CRE-2 element. Instead, as determined by antibody supershift analysis, we detect CRE-binding protein (CREB)-like proteins binding to CRE-2 in both basal and haloperidol-stimulated conditions. Finafly, we show that haloperidol induces CREB protein phosphorylation in striatum.

Membrane depolarization is a critical component of neural signaling; in recent years there also has been a great deal of evidence that membrane depolarization can regulate neural gene expression. Therefore, excitatory neurotransmission may be an important mechanism of neural plasticity. We have investigated the intracellular pathways and DNA regulatory elements through which membrane depolarization activates expression of the neural gene encoding human proenkephalin. In PC12 and C6-glioma cells, depolarization-induced expression of a transfected proenkephalin fusion gene was proportional to extracellular calcium concentration and was inhibited by verapamil. Activation of the gene by KCl-induced depolarization or the calcium ionophore A23187 was dependent upon and synergistic with cAMP in PC12 and C6-glioma cells, but neither depolarization nor treatment with A23187 affected cAMP levels. Trifluoperazine and W7 inhibited depolarization-induced gene expression but did not affect expression induced by the adenylyl cyclase activator forskolin. At the level of the DNA, depolarization- induced activation is conferred on the proenkephalin gene by a previously characterized cAMP-inducible enhancer. Multiple copies of a single component element of that enhancer, containing the CGTCA sequence motif characteristic of cAMP regulatory elements, can reconstitute the entire repertoire of responses to both cAMP and depolarization. These data suggest a model in which membrane depolarization activates gene expression through a calcium-dependent pathway, potentially involving calmodulin, and in which the transcriptional responses to both cAMP and calcium are transduced by the same DNA element.

We demonstrate that JunD, a component of the AP-1 transcription factor complex, activates transcription of the human proenkephalin gene in a fashion that is completely dependent upon the cAMP-dependent protein kinase, protein kinase A. Activation of proenkephalin transcription by JunD is dependent upon a previously characterized cAMP-, phorbol ester-, and Ca(2+)-inducible enhancer, and JunD is shown to bind the enhancer as a homodimer. Another component of the AP-1 transcription complex, JunB, is shown to inhibit activation mediated by JunD. As a homodimer JunB is unable to bind the enhancer; however in the presence of c-Fos, high-affinity binding is observed. Furthermore, JunD is shown to activate transcription of genes linked to both cAMP and phorbol ester response elements in a protein kinase A-dependent fashion, further blurring the distinction between these response elements. These results demonstrate that the transcriptional activity of an AP-1-related protein is regulated by the cAMP-dependent second-messenger pathway and suggest that JunD and other AP-1-related proteins may play an important role in the regulation of gene expression by cAMP-dependent intracellular signaling pathways.

In the preceding paper (Kobierski et al: J. Comp. Neurol. 266:1-15, '87) FMRFamidelike immunoreactivity (FLI) was localized to specific cells and processes in the nervous system of the lobster Homarus americanus. In an effort to establish a role for this material we have purified and characterized a variety of immunoreactive peptides that can be extracted from the secretory pericardial organs. By using gel-filtration chromatography and three different HPLC systems, it has been established that little or no authentic FMRFamide is present. Of the major immunoreactive components two peptides were purified in sufficient quantity for microsequence analysis and have been tentatively identified as the octapeptides Ser-Asp-Arg-Asn-Phe-Leu-Arg-Phe-amide (FLI 3) and Thr-Asn-Arg-Asn-Phe-Leu-Arg-Phe-amide (FLI 4). Both of these are novel neuropeptides with some sequence homology to the previously described FMRFamide family. The pericardial organs release FLI when depolarized with 100 mM K+ in the presence of calcium. Between 75 and 80% of this release is accounted for by FLI 3 and FLI 4. One of these peptides (FLI 4) has been synthesized and shown to cochromatograph with the endogenous immunoreactive material. Preliminary studies show that this peptide can act as a modulator of exoskeletal and cardiac neuromuscular junctions.

The distribution of FMRFamidelike peptides was studied in the nervous system of the lobster Homarus americanus by using immunocytochemical and radioimmunological techniques. By radioimmunoassay FMRFamidelike immunoreactivity (FLI) was found in low levels (ca. 1 pmol/mg protein) throughout the ventral nerve cord and in much higher amounts (60-100 pmol/mg protein) in the neurosecretory pericardial organs. Immunocytochemical studies showed FLI in approximately 300-350 cell bodies, and in distinct neuropil regions, neuronal fiber tracts, and varicose endings. Specificity of the immunostaining was tested by preabsorbing the antiserum with FMRFamide, with peptides having similar carboxyl termini to FMRFamide (Met-enkephalin-Arg-Phe, Phe-Met-Arg-Tyr-amide), with several amidated peptides (alpha-melanocyte-stimulating hormone, substance P, oxytocin), and with proctolin, a peptide found widely distributed in the lobster nervous system. Of these substances, only FMRFamide blocked the staining. In addition to the pericardial organs, significant levels of FLI were found in neurosecretory regions associated with thoracic second roots and in the connective tissue sheath that surrounds the ventral nerve cord. In all three regions, immunocytochemical studies showed the FLI to be localized to fine fibers and associated terminal varicosities lying close to the surface of the tissue, with no obvious target in their immediate vicinity. When examined at the ultrastructural level, the immunoreactive varicosities of the thoracic second roots and of the ventral nerve cord sheaths were found a few microns from the surface of the tissue and contained electron-dense granules. In the immunoreactive nerve cord sheath endings, in addition to the large, dense granules, small, clear vesicles were found. The appearance and location of these terminals suggest a neurohormonal role for FMRFamidelike peptides in lobsters. The observation that low levels of FLI are found in the hemolymph supports this suggestion. In addition, the localization of FLI to particular neuronal somata, fiber tracts, and neuropil regions suggests possible functional roles for these peptides in (1) integration of visual and olfactory information, (2) function of the anterior and posterior gut, and (3) the control of exoskeletal muscles.