Plasticity is a property of the nervous system that allows it to modify its response to an altered input. This capacity for change suggests that there are molecular mechanisms in neurons that can couple stimuli to long-term alterations in phenotype. Neuronal excitation elicits rapid transcriptional activation of several immediate-early genes, for example c-fos, c-jun and zif268. Many immediate-early genes encode transcription factors that control expression of downstream genes whose products are believed to bring about long-term plastic changes. Here we use a highly sensitive differential complementary DNA cloning procedure to identify genes that may participate in long-term plasticity. We cloned 52 cDNAs of genes induced by the glutamate analogue kainate in the hippocampus dentate gyrus. The number of these candidate plasticity-related genes (CPGs) is estimated to be 500-1,000. One of the cloned CPGs (16C8), encoding a protease inhibitor, is induced by a stimulus producing long-term potentiation and during dentate gyrus development; a second, cpg1, is dependent on activation of the NMDA (N-methyl-D-aspartate) receptor for induction and encodes a new small, dentate-gyrus-specific protein. Seventeen of the cloned CPGs encode known proteins, including six suggesting that strong neuronal activation leads to de novo synthesis of vesicular and other synaptic components.

Neural activity and neurotrophins induce synaptic remodeling in part by altering gene expression. A cDNA encoding a glycosylphoshatidylinositol-anchored protein was identified by screening for hippocampal genes that are induced by neural activity. This molecule, named neuritin, is expressed in postmitotic-differentiating neurons of the developing nervous system and neuronal structures associated with plasticity in the adult. Neuritin message is induced by neuronal activity and by the activity-regulated neurotrophins BDNF and NT-3. Purified recombinant neuritin promotes neurite outgrowth and arborization in primary embryonic hippocampal and cortical cultures. These data implicate neuritin as a downstream effector of activity-induced neurite outgrowth.

Complementary DNA cloning of the transcripts of the Drosophila clock gene period reveals three distinct transcripts. These result from unusual splicing pathways, one involving a CG 3' splice site and one resulting in the use of two different reading frames in one exon, and they predict three separate proteins. Two of the cloned cDNAs can restore clock function to mutant arrhythmic flies.

Long-term plasticity of the central nervous system (CNS) involves induction of a set of genes whose identity is incompletely characterized. To identify candidate plasticity-related genes (CPGs), we conducted an exhaustive screen for genes that undergo induction or downregulation in the hippocampus dentate gyrus (DG) following animal treatment with the potent glutamate analog, kainate. The screen yielded 362 upregulated CPGs and 41 downregulated transcripts (dCPGs). Of these, 66 CPGs and 5 dCPGs are known genes that encode for a variety of signal transduction proteins, transcription factors, and structural proteins. Seven novel CPGs predict the following putative functions: cpg2--a dystrophin-like cytoskeletal protein; cpg4--a heat-shock protein: cpg16--a protein kinase; cpg20--a transcription factor; cpg21--a dual-specificity MAP-kinase phosphatase; and cpg30 and cpg38--two new seven-transmembrane domain receptors. Experiments performed in vitro and with cultured hippocampal cells confirmed the ability of the cpg-21 product to inactivate the MAP-kinase. To test relevance to neural plasticity, 66 CPGs were tested for induction by stimuli producing long-term potentiation (LTP). Approximately one-fourth of the genes examined were upregulated by LTP. These results indicate that an extensive genetic response is induced in mammalian brain after glutamate receptor activation, and imply that a significant proportion of this activity is coinduced by LTP. Based on the identified CPGs, it is conceivable that multiple cellular mechanisms underlie long-term plasticity of the nervous system.

The pero1 and the pers mutations in Drosophila melanogaster, which seem to eliminate or speed up, respectively, the clocks underlying biological rhythmicity, were mapped to single nucleotides. Chimeric DNA fragments consisting of well-defined wild-type plus mutant DNA subsegments were constructed, introduced into flies by germ-line transformation, and assayed for biological activity. These experiments localized both perol and pers to a 1.7-kilobase DNA fragment that is mostly coding DNA. Sequencing of this subsegment from each mutant showed that per"1 is completely accounted for by a nonsense mutation in the third coding exon of a 4.5-kilobase RNA transcribed from this locus. The pers mutation is also a single nucleotide substitution, in the fourth coding exon, which results in a serine-to-asparagine substitution in the per gene protein product. The functional significance of these changes is discussed with reference to the phenotypes of the two mutations.The period gene of Drosophila melanogaster is defined by some of the most interesting behavioral mutations isolated to date. These mutations affect biological rhythms either by eliminating periodic eclosion or behavior or by causing dramatic changes in rhythmic phenotypes (for review see refs. 1 and 2). The original per mutations were induced by Konopka and Benzer (3). Two of the mutants were per' and perL, isolated with respect to circadian eclosion rhythms whose periods are about 5 hr shorter or longer, respectively, than the normal 24 hr. The per' mutant was also recovered (3), based on aperiodic eclosion. This mutation is now called per"l, because there are currently four independently isolated "arrhythmic" mutations at this locus (refs. 4 and 5; R. J. Konopka and M.H., unpublished results).Detailed genetic and phenotypic analyses of these "clock" mutants suggested that per'1 may be a null mutation, behaving like a deletion of this apparently nonvital locus within segment 3B1-2 of the X chromosome (6, 7). Manipulations of the dosage of the wild-type allele and the pers allele suggested that the increased per activity seemingly associated with pers is caused by a qualitative change in the activity of the gene product (4,8).A 6-kilobase (kb) subset of the per region (9, 10) appears to be necessary for per' activity-i.e., rescues per'-induced arrhythmicity of (i) circadian locomotor activity (5, 11, 12), (ii) eclosion (12), and (iii) variations in the male courtship "singing" behavior (5, 12). This portion of the per region corresponds to a 4.5-kb RNA species (9, 10), allowing one to focus on its informational content (13-15).The results we present here include a determination of the precise sites of the mutations in the original arrhythmic and short-period variants. These experiments were undertaken to address some of the issues raised above and the role of the gene in the development (e.g., refs. 16 and 17) and the function (e.g., refs. 15 and 18) of biological clocks in Drosophila. Thus, we set out to determine ifperol is in fact a null mutant and t...