Published papers

Full-length cDNA for human ciliary neurotrophic factor (CNTF) was isolated from a human sciatic nerve cDNA library. Sequence analysis revealed that the longest cDNA was comprised of a 48-bp 5′-untranslated region, a 600-bp coding region and a 1207-bp 3′-untranslated region containing four ATTA pentamer motifs and a polyadenylation sequence. The transcription starting point was assigned at 81 bp upstream of the initiation methionine by 5′ RACE analysis. Using the cDNA and genomic DNA fragment including the entire intron region as mixed probes, the human CNTF gene was localized to the long arm of chromosome 11 at region q12 by fluorescence in situ hybridization.

We report a null mutation in the human ciliary neurotrophic factor gene (CNTF). The mutated allele shows a G to A transition producing a new splice acceptor site and the resulting mRNA species codes for an aberrant protein. Analysis of tissue samples and transfection of CNTF minigenes into cultured cells demonstrates that the mutated allele expresses only the mutated mRNA species. In 391 Japanese people tested, 61.9% were normal homozygotes, 35.8% heterozygotes and 2.3% mutant homozygotes. The distribution of the three genotypes is similar in healthy and neurological disease subjects, indicating that human CNTF deficiency is not causally related to neurological diseases.

Stably transfected cells expressing mouse choline acetyltransferase (ChAT) cDNA were established, and the synthesis and release of acetylcholine (ACh) were examined. A cDNA clone coding for mouse ChAT was inserted into an expression vector (pEF321) containing a promoter for human elongation factor 1α to construct pEFmChAT. Neuronal (NG108-15, NS20Y, N1E115, and Neuro2A) and nonneuronal cell lines (L cells and NIH3T3) were transfected with pEFmChAT, and the cell lines that stably expressed high ChAT activity were selected. These cells expressed the 66-kDa ChAT protein and accumulated ACh mostly in the cytosol. The concentration of intracellular ACh in the cells increased upon raising the choline level in the medium. The cells continuously released ACh in a Ca2+-independent fashion. Neither high K+ nor calcium ionophore stimulated release of ACh from the cells.

A cDNA library directed by a specific primer was constructed from the rat spinal cord and screened with 32P-labeled rat choline acetyltransferase cDNA which was recently isolated in this laboratory. Sequence analysis of 29 clones indicated that there are four types of cDNA (R1-, R2-, N1- and M-types). The nucleotide sequences in these cDNAs were identical in the coding region and the first 38 bp of the 5′-noncoding region, but differed in the 5′-noncoding region upstream of -38 bp. The R1-type was identical to the cDNA previously cloned from the rat spinal cord. The M and N1-type cDNAs both had sequences homologous to that of the cDNA previously obtained from the mouse spinal cord. Polymerase chain reaction analysis confirmed the presence of these 4 types of mRNA and found another type (N2-type) of transcript. The numbers of cDNA clones isolated and the relative amounts of polymerase chain reaction products for each type of mRNA suggested that the most abundant transcript was M-type. Sequencing of the genomic clone containing the 5′-region of choline acetyltransferase mRNA revealed that these five types of mRNA species were transcribed from three different promoter regions and produced by differential splicing of the 5′-noncoding exons.

The effect of cyclic AMP on the gene expression of cholinc acetyltransferase (ChAT) was studied in NG108-15, mouse neuroblastoma and rat glioma hybrid cell lines. Addition of dibutyryl cyclic AMP to the culture medium increased both the ChAT mRNA level and ChAT activity twofold. Polymerase chain reaction analysis of the ChAT mRNA indicated that, among the multiple mRNA species, M-type mRNA was transcribed most efficiently, with or without the addition of dibutyryl cyclic AMP. The 5́ region of the mouse ChAT gene was ligated to the bacterial chloramphenicol acetyltransferase gene, and the expression of chloramphenicol acetyltransferase activity was determined by transaction analysis. Cyclic AMP derivatives enhanced the reporter gene expression in both transiently and stably transfected cells. DNA deletion analysis indicated that the intron region downstream of the M-type exon is necessary for the cyclic AMP responsiveness, and that cyclic AMP derivatives increase ChAT gene transcription mainly from M-type promoter. These results suggest that a cis-acting DNA element that confers the cyclic AMP responsiveness of the ChAT gene is present in the intron downstream of the M-type exon.

Seven types of mRNA for choline acetyltransferase that differ in the 5′-noncoding region were identified in the mouse spinal cord by cDNA cloning and polymerase chain reaction. Among these transcripts, the M-type mRNA corresponding to the previously cloned mouse cDNA was most abundant in the spinal cord of mouse. A mouse genomic DNA clone containing the 5′-region of choline acetyltransferase mRNA was isolated and sequenced. Comparison of the sequences between the cDNAs and the genomic DNA revealed that the different mRNA species were transcribed from different promoter regions and produced by differential splicing. Two murine cholinergic cell lines, NS20Y and NG108-15, were shown to express the M-type mRNA almost exclusively, and were therefore used to study transcription of M-type mRNA. Fragments of the 5′-region of choline acetyltransferase gene were ligated with chloramphenicol acetyltransferase reporter gene and introduced into cultured cells. The fragment from -2752 to +46, which contained the M-type exon, a TATA-box like element upstream of the M-type exon, and the downstream intron, induced a significant expression of CAT activity in neuronal but not in non-neuronal cell lines. This result indicates that this region of choline acetyltransferase gene contains elements that regulate neuron-specific expression of choline acetyltransferase activity. However, there was no parallel correlation between reporter gene expression in the transfected cells and intrinsic choline acetyltransferase activity in these neuronal cell lines. Possible mechanisms that would explain this observation are discussed.

Functional coupling between μ-opioid receptors and GTP-binding regulatory proteins (G proteins) was investigated in reconstituted membranes of the guinea pig striatum. Selective μ-opioid agonists stimulated low-Km GTPase in striatal membranes, in a Na+-dependent manner. The same μ-opioid agonist {[D-Ala2,N-Me-Phe4,Gly5-ol]-enkephaIin (DAGO)} caused no stimulation when the membranes were exposed to islet-activating protein (IAP; pertussis toxin). There was also no DAGO stimulation in preparations pretreated with a lower concentration (5 μM) of N-ethylmaleimide (NEM), which abolished the ADP-ribosylation of purified Gi( (the G protein that mediates inhibition of adenylate cyclase) and Go (a G protein of unknown function purified from bovine brain) by IAP. In addition, as the NEM treatment caused no change in the μ-agonist binding, NEM could probably substitute for IAP in inactivating native G proteins, without exhibiting effects on the receptor binding in mem-branes. The μ-agonist stimulation of low-Km GTPase activity in NEM-treated membranes was recovered by reconstitution with purified Gi or Go. The μ-agonist stimulation of low-Km GTPase was additive when Gi and Go were simultaneously reconstituted in NEM-treated membranes in amounts of 0.5 pmol/assay, which was required for maximal recovery, in either reconstitution experiment. The present findings provide the first evidence that the μ-opioid receptor may exist in at least two different forms, separately coupled to G1 or Go.

In synaptic membranes of guinea pig cerebellum, in which preparations U-50, 488H, a κ-opioid agonist inhibits G-protein activity, the same κ-agonist at 100 nM inhibited the phospholipase C (PLC) activity stimulated by 100 μM GTP in the presence of 10 μM Ca2+. The present study provides a unique hypothetical mechanism that the receptor-mediated inhibition of PLC is mediated by inhibition of G-protein activity.

We attempted to identify the kyotorphin receptor and the post receptor mechanisms mediated by GTP-binding proteins (G-proteins), using reconstitution techniques. The specific binding of [3H]kyotorphin in rat brain membranes was composed of high affinity (K(d) = 0.34 nM) and low affinity (K(d) = 9.07 nM) binding. As the high affinity binding disappeared in the presence of guanosine 5'-O-(3-thiotriphosphate) and MgCl2, we investigated the kyotorphin receptor-mediated changes in membrane G-protein activity by measuring low K(m) GTPase activity. Kyotorphin produced a stimulation of low K(m) GTPase, and this stimulation was antagonized by Leu-Arg, a synthetic dipeptide which showed a potent displacement of [3H]kyotorphin binding, yet in itself had no effect on the low K(m) GTPase. The kyotorphin stimulation of low K(m) GTPase was abolished by pretreating membranes with islet-activating protein, pertussis toxin, and was recovered by reconstitution with purified G-protein, G(i), but not with G(o). Similar evidence of selective coupling of kyotorphin receptor to G(i) was obtained with the phospholipase C assay. Kyotorphin-induced stimulation of phospholipase C was also abolished by islet-activating protein-treatment and recovered by reconstitution with G(i) but not with G(o). These findings indicate that specific high and low affinity kyotorphin receptors exist in the rat brain and that the kyotorphin receptor is functionally coupled to stimulation of phospholipase C, through G(i). This study provides the first evidence of a selective involvement of G(i) in the receptor-mediated activation of phospholipase C.

DAGO, a specific opioid μ-agonists stimulated the low Km GTPase activity of membranes of the guinea-pig striatum at 1-1000 nM and this effect was antagonized by naloxone 10-100 nM. On the other hand, U-50,488H, a specific opioid κ-agonist inhibited the low Km GTPase activity of membranes of the guinea-pig cerebellum at 1-100 nM. As this effect was antagonized by MR2266 but not by naloxone, the findings suggest that opioid μ- and κ-receptor stimulation is probably linked to the stimulation and inhibition of low Km GTPase, respectively.

Opioid μ-receptors were solubilized from rat brains with 0.5% Triton X-100 in the presence of 0.32 M sucrose and then purified with 6-succinyl morphine-Affi-Gel 102 column. Purified materials showed a major band with mol.wt. 58,000 Da and a minor band with mol.wt. 41,000 Da. Affinity cross linking with [3H]DAGO, a selective μ-opioid agonist, to the neural membranes and purified materials revealed that opioid μ-binding protein has a mol.wt. of 58,000 Da, a value totally different from that of δ- (mol.wt. 18,500 Da) and κ- (mol.wt. 36,600 Da) binding protein.

Choline acetyltransferase (ChAT) is a specific phenotypic marker of cholinergic neurons. Previous reports showed that different upstream regions of the ChAT gene are necessary for cell type-specific expression of reporter genes in cholinergic cell lines. The identity of the mouse ChAT promoter region controlling the establishment, maintenance, and plasticity of the cholinergic phenotype in vivo is not known. We characterized a promoter region of the mouse ChAT gene in transgenic mice, using β-galactosidase (LacZ) as a reporter gene. A 3,402-bp segment from the 5'-untranslated region of the mouse ChAT gene (from -3,356 to +46, +1 being the translation initiation site) was sufficient to direct the expression of LacZ to selected neurons of the nervous system; however, it did not provide complete cholinergic specificity. A larger fragment (6,417 bp, from -6,371 to +46) of this region contains the requisite regulatory elements that restrict expression of the LacZ reporter gene only in cholinergic neurons of transgenic mice. This 6.4-kb DNA fragment encompasses 633 bp of the 5'- flanking region of the mouse vesicular acetylcholine transporter (VAChT), the entire open reading frame of the VAChT gene, contained within the first intron of the ChAT gene, and sequences upstream of the start coding sequences of the ChAT gene. This promoter will allow targeting of specific gene products to cholinergic neurons to evaluate the mechanisms of diseases characterized by dysfunction of cholinergic neurons and will be valuable in design strategies to correct those disorders.

Classical neurotransmitters such as γ-aminobutyric acid and glutamate are released from synaptic nerve terminals by exocytosis of synaptic vesicles. PC12 cells also have SSVs capable of storing acetylcholine (ACh). A novel method to examine the effect of transient transfection of any gene of interest on the exocytosis of SSVs was developed. The transfection of choline acetyltransferase (ChAT) into PC12 cells which have lost ACh synthesizing activity resulted in the accumulation of a substantial amount of ACh. Synthesized ACh was released in Ca2+-dependent manner. Release was thought to occur by an exocytosis of SSVs because: (1) release was abolished by treating the cells with vesamicol, a specific inhibitor of the vesicular ACh transporter (VAChT) localizing specifically in SSVs; and (2) the release was further increased by cotransfecting rat VAChT with the ChAT. By means of this method, we showed that overexpression of complexin I or II with ChAT markedly suppressed high-K+-dependent ACh release of SSVs.

Both muscarinic and nicotinic acetylcholine (ACh) receptors are known to be present on the surface of lymphocytes. We have shown that variable amounts of ACh are detectable in the blood of various mammals including humans, and a major portion of blood ACh is localized in circulating mononuclear leukocytes in humans. In order to investigate which types of blood cell are the source of ACh in human blood, expression of mRNA for choline acetyltransferase (CHAT, EC 2.3.1.6), which catalyzes ACh synthesis, was analyzed using human leukemic cell lines as models of lymphocytes and the reverse transcription- polymerase chain reaction (RT-PCR) method. We observed that mRNA for the same ChAT as that in the nervous system is expressed constitutively in all the T- cell lines tested, but not in B-, pre-lymphoma or monocytic cell lines. Furthermore, only T-cell lines showed high ACh-synthesizing activities and intracellular ACh contents. These results suggest that the major portion of ACh in the circulating blood originates from T-lymphocytes.

The expression of vesicular acetylcholine transporter (VAChT) was examined immunohistochemically in the cholinergic synapses on the spinal motor neuron of the patient with sporadic amyotrophic lateral sclerosis (SALS). VAChT immunoreactive synapses were depleted on surviving motor neurons in SALS, while synaptophysin immunoreactivity was undiminished on the same neurons. This discrepancy suggests that in SALS, loss of cholinergic input on lower motor neurons is an early event, and may be part of the cause of death of those motor neurons.

We used a sensitive and specific radioimmunoassay for acetylcholine (ACh), and detected significant amounts of ACh in the blood of various mammals, including humans. About 60% of human blood ACh was localized in mononuclear leukocytes. Human leukemic T-cell lines, used as T-lymphocyte models, contained both ACh and choline acetyltransferase (ChAT) activity. Furthermore, ChAT mRNA and protein were detected in the T-cell line MOLT-3. Phytohemagglutinin, a T-cell activator, increased both synthesis and release of ACh by MOLT-3 cells. Muscarinic receptor subtype mRNA expression was confirmed in various T-cell lines. These findings indicate that ACh synthesized by ChAT in T-lymphocytes acts on the muscarinic receptors on lymphocytes in autocrine and/or paracrine pathways and suggest that ACh in blood functions as a modulator of T-cell-dependent immune responses.

The induction of mRNA for choline acetyltransferase (CHAT), which catalyzes acetylcholine (ACh) synthesis was investigated in human mononuclear leukocytes (MNL) stimulated by phytohemagglutinin (PHA), a T-cell activator, using the reverse transcription-polymerase chain reaction. Stimulation of MNL by PHA induced the expression of ChAT mRNA, and potentiated ACh synthesis. ChAT mRNA induction required more time than the induction of interleukin-2 mRNA. Expression of the gene encoding the vesicular ACh transporter, which mediates ACh transport in cholinergic neurons, was not observed in PHA- stimulated MNL, suggesting that the mechanisms controlling ACh release from T-lymphocytes differ from those in cholinergic neurons. These findings demonstrate that activation of T-lymphocytes up-regulates ACh synthesis in the blood, and suggest that ACh plays an important role as a neuroimmunomodulator besides its role as a neurotransmitter.

The organization of the mouse choline acetyltransferase (ChAT) gene has been previously analyzed. Here we show that the first intron of the mouse ChAT gene contains an uninterrupted open reading frame. It is in the same transcriptional orientation as ChAT and encodes the vesicular acetylcholine (ACh) transporter (VAChT), the protein responsible for the translocation of cytoplasmic ACh into synaptic vesicles. The sequence of this transporter is very similar to the VAChT from rat and human (99% and 95% identity, respectively). Reverse transcription-polymerase chain reaction (RT-PCR) analysis showed expression of mouse VAChT mRNA in spinal cord, brain (excluding the cerebellum) and brain stem, but not in peripheral tissues such as liver and kidney. Transgenic mouse analysis revealed that the 5'-flanking region of the mouse ChAT gene encompasses regulatory elements that allowed elevated expression of VAChT in the cholinergic system of transgenic mice.

Choline acetyltransferase (ChAT, EC 2.3.1.6) is the biosynthetic enzyme for acetylcholine. We have previously shown that multiple ChAT mRNA species with different 5'-noncoding regions are expressed in the rat and mouse. However, the diversity of ChAT mRNA species in human has not completely been elucidated. In this work N1- and N2-type ChAT cDNAs were cloned from a human brain cDNA library and the N-exon located in the human ChAT gene. Polymerase chain reaction analysis indicates that four species of ChAT mRNAs (R-, N1-, N2- and M-types) are produced in human brain and spinal cord. In all human transcripts, the ATG initiation codon in the rat, mouse and pig was replaced by ACG, which does not serve as an initiation codon for translation. In vitro translation and mammalian expression analyses revealed that N1-, N2- and R-type mRNAs give rise to a single 69 kDa enzyme, while M-type mRNA produces both 82 and 69 kDa enzymes. The translation efficiency of M-type mRNA was lower than that of the other mRNA species. Moreover, the translation efficiency of human ChAT mRNAs was considerably lower than that of rat ChAT mRNA, suggesting that the ATG codons for human ChAT are unfavorable for translation initiation compared with the initiation codon for rat ChAT. These results provide rational explanations for the previous reports that human ChAT protein purified from the brain acid placenta had 66-70 kDa molecular mass, and that ChAT activity in a single motor neuron of human was far lower than that of other vertebrates. Sequencing of monkey ChAT gene showed that the initiation ATG in rodent ChAT was also replaced by ACA in the monkey.

Choline acetyltransferase (ChAT) is a specific phenotypic marker of cholinergic neurons. Previous reports showed that different upstream regions of the ChAT gene are necessary for cell-type specific expression of reporter genes in cholinergic cell lines. The identity of the mouse ChAT promoter region controlling the establishment, maintenance, and plasticity of the cholinergic phenotype in vivo is not known. We characterized the promoter region of the mouse ChAT gene in transgenic mice, using β-galactosidase (β-gal) as a reporter. A 3401-bp segment from the 5′-untranslated region of the mouse ChAT gene was sufficient to direct the expression of β-gal to selected neurons of the nervous system (NS); it did not provide cholinergic specificity A larger fragment (∼6 Kb) of this region encompassing most of the coding sequence for the putative mouse vesicular acetylcholine transporter (VAChT) contained within the first intron of the ChAT gene, directs correct developmental and tissue-specific expression to cholinergic cells in transgenic mice. This fragment targeted β-gal expression only to cholinergic regions of the NS of transgenic mice, including basal forebrain, striatum, cortex, brain stem, and exclusively to α-motor neurons in the ventral horn of the spinal cord. This ChAT promoter will allow targeting of specific gene products to cholinergic neurons, in order to develop animal models for human diseases where injury and degeneration of chotinergic neurons is involved.