Differentiation induces up-regulation of plasma membrane Ca(2+)-ATPase and concomitant increase in Ca(2+) efflux in human neuroblastoma cell line IMR-32.

Precise regulation of intracellular Ca(2+) concentration ([Ca(2+)](i)) is achieved by the coordinated function of Ca(2+) channels and Ca(2+) buffers. Neuronal differentiation induces up-regulation of Ca(2+) channels. However, little is known about the effects of differentiation on the expression of the plasma membrane Ca(2+)-ATPase (PMCA), the principal Ca(2+) extrusion mechanism in neurons. In this study, we examined the regulation of PMCA expression during differentiation of the human neuroblastoma cell line IMR-32. [Ca(2+)](i) was monitored in single cells using indo-1 microfluorimetry. When the Ca(2+)-ATPase of the endoplasmic reticulum was blocked by cyclopiazonic acid, [Ca(2+)](i) recovery after small depolarization-induced Ca(2+) loads was governed primarily by PMCAs. [Ca(2+)](i) returned to baseline by a process described by a monoexponential function in undifferentiated cells (tau = 52 +/- 4 s; n = 25). After differentiation for 12-16 days, the [Ca(2+)](i) recovery rate increased by more than threefold (tau = 17 +/- 1 s; n = 31). Western blots showed a pronounced increase in expression of three major PMCA isoforms in IMR-32 cells during differentiation, including PMCA2, PMCA3 and PMCA4. These results demonstrate up-regulation of PMCAs on the functional and protein level during neuronal differentiation in vitro. Parallel amplification of Ca(2+) influx and efflux pathways may enable differentiated neurons to precisely localize Ca(2+) signals in time and space.

The calmodulin multigene family as a unique case of genetic redundancy: multiple levels of regulation to provide spatial and temporal control of calmodulin pools?

Calmodulin (CaM) is a ubiquitous, highly conserved calcium sensor protein involved in the regulation of a wide variety of cellular events. In vertebrates, an identical CaM protein is encoded by a family of non-allelic genes, raising questions concerning the evolutionary pressure responsible for the maintenance of this apparently redundant family. Here we review the evidence that the control of the spatial and temporal availability of CaM may require multiple regulatory levels to ensure the proper localization, maintenance and size of intracellular CaM pools. Differential transcription of the CaM genes provides one level of regulation to meet tissue-specific, developmental and cell-specific needs for altered CaM levels. Post-transcriptional regulation occurs at the level of mRNA stability, perhaps dependent on alternative polyadenylation and differences in the untranslated sequences of the multiple gene transcripts. Recent evidence indicates that trafficking of specific CaM mRNAs may occur to specialized cellular locales such as the dendrites of neurons. This could allow local CaM synthesis and thereby help generate local pools of CaM. Local CaM activity may be further regulated by post-translational mechanisms such as phosphorylation or storage of CaM in a 'masked' form. The spatial resolution of CaM activity is enhanced by the limited free diffusion of CaM combined with differential affinity for and availability of target proteins. Preserving multiple CaM genes with divergent noncoding sequences may be necessary in complex organisms to ensure that the many CaM-dependent processes occur with the requisite spatial and temporal resolution. Transgenic mouse models and studies on mice carrying single and double gene 'knockouts' promise to shed further light on the role of specificity versus redundancy in the evolutionary maintenance of the vertebrate CaM multigene family.

Characterization of the human CALM2 calmodulin gene and comparison of the transcriptional activity of CALM1, CALM2 and CALM3.

Human calmodulin is encoded by three genes CALM1, CALM2 and CALM3 located on different chromosomes. To complete the characterization of this family, the exon-intron structure of CALM2 was solved by a combination of genomic DNA library screening and genomic PCR amplification. Intron interruptions were found at identical positions in human CALM2 as in CALM1 and CALM3; however, the overall size of CALM2 (16 kb) was almost twice that of the other two human CALM genes. Over 1 kb of the 5' flanking sequence of human CALM2 were determined, revealing the presence of a TATA-like sequence 27 nucleotides upstream of the transcriptional start site and several conserved sequence elements possibly involved in the regulation of this gene. To determine if differential transcriptional activity plays a major role in regulating cellular calmodulin levels, we directly measured and compared the mRNA abundance and transcriptional activity of the three CALM genes in proliferating human teratoma cells. CALM3 was at least 5-fold more actively transcribed than CALM1 or CALM2. CALM transcriptional activity agreed well with the mRNA abundance profile in the teratoma cells. In transient transfections using luciferase reporter genes driven by 1 kb of the 5' flanking DNA of the three CALM genes, the promoter activity correlated with the endogenous CALM transcriptional activity, but only when the 5' untranslated regions were included in the constructs. We conclude that the CALM gene family is differentially active at the transcriptional level in teratoma cells and that the 5' untranslated regions are necessary to recover full promoter activation.

Minimum CAG repeat in the human calmodulin-1 gene 5' untranslated region is required for full expression.

The human calmodulin-1 gene (hCALM1) contains a (CAG)7 repeat in its 5'-untranslated region (5'-UTR). We found this repeat to be stable and nonpolymorphic in the human population. To determine whether the repeat region affects hCALM1 expression and whether repeat expansions to numbers known to be associated with disease in other genes may alter expression, we tested luciferase reporter genes driven by the hCALM1 promoter and 5'-UTR containing 0, 7 (wild-type), 20, and 45 CAG repeats in human NT2/D1 teratoma cells. Interestingly, the repeat deletion, (CAG)0, decreased expression by 45%, while repeat expansions to (CAG)20 and (CAG)45, or the insertion of a scrambled (C,A,G)7 sequence did not alter gene expression. These data indicate (1) that the endogenous repeat element is required for full expression of hCALM1, and (2) that some triplet repeat expansions in the 5'-UTR of protein-coding genes may be well tolerated and even optimize gene expression.

Tyrosine kinase activity may be necessary but is not sufficient for c-erbB1-mediated tissue-specific tumorigenicity.

Expression of mutant avian c-erbB1 genes results in tissue-specific transformation in chickens. Site-directed mutagenesis was used to generate kinase-defective mutants of several tissue-specific v-erbB transforming mutants by replacement of the ATP-binding lysine residue in the kinase domain with an arginine residue. These kinase-defective v-erbB mutants were analyzed for their in vitro and in vivo transforming potentials. Specifically, kinase-defective mutants of erythroleukemogenic, hemangioma-inducing, and sarcomagenic v-erbB genes were assessed for their oncogenic potential. In vitro transformation potential was assessed by soft-agar colony formation in primary cultures of chick embryo fibroblasts (CEF). In vivo transformation potential was determined by infection of 1-day-old line 0 chicks with concentrated recombinant retrovirus and then monitoring of birds for tumor formation. These transformation assays demonstrate that kinase activity is absolutely essential for transformation by tissue-specific transforming mutants of the avian c-erbB1 gene. Since all of the tissue-specific v-erbB mutants characterized to date exhibit tyrosine kinase activity in vitro but do not transform all tissues in which they are expressed, we conclude that v-erbB-associated tyrosine kinase activity may be necessary but is not sufficient to induce tumor formation.

Regulation of phosphatidylethanolamine methyltransferase and phospholipid methyltransferase by phospholipid precursors in Saccharomyces cerevisiae.

Phosphatidylethanolamine methyltransferase (PEMT) and phospholipid methyltransferase (PLMT), which are encoded by the CHO2 and OPI3 genes, respectively, catalyze the three-step methylation of phosphatidylethanolamine to phosphatidylcholine in Saccharomyces cerevisiae. Regulation of PEMT and PLMT as well as CHO2 mRNA and OPI3 mRNA abundance was examined in S. cerevisiae cells supplemented with phospholipid precursors. The addition of choline to inositol-containing growth medium repressed the levels of CHO2 mRNA and OPI3 mRNA abundance in wild-type cells. The major effect on the levels of the CHO2 mRNA and OPI3 mRNA occurred in response to inositol. Regulation was also examined in cho2 and opi3 mutants, which are defective in PEMT and PLMT activities, respectively. These mutants can synthesize phosphatidylcholine when they are supplemented with choline by the CDP-choline-based pathway but they are not auxotrophic for choline. CHO2 mRNA and OPI3 mRNA were regulated by inositol plus choline in opi3 and cho2 mutants, respectively. However, there was no regulation in response to inositol when the mutants were not supplemented with choline. This analysis showed that the regulation of CHO2 mRNA and OPI3 mRNA abundance by inositol required phosphatidylcholine synthesis by the CDP-choline-based pathway. The regulation of CHO2 mRNA and OPI3 mRNA abundance generally correlated with the activities of PEMT and PLMT, respectively. CDP-diacylglycerol synthase and phosphatidylserine synthase, which are regulated by inositol in wild-type cells, were examined in the cho2 and opi3 mutants. Phosphatidylcholine synthesis was not required for the regulation of CDP-diacylglycerol synthase and phosphatidylserine synthase by inositol.