Anticonvulsants-induced chorea: a role for pharmacodynamic drug interaction?

Chorea is a rare side effect of anticonvulsants. We describe three patients who developed chorea secondary to anticonvulsant combination use. A mechanism to explain this finding is proposed. After identification of an index case with anticonvulsant-induced chorea, we reviewed the electronic data base records for all patients with seizures followed in the epilepsy clinics at our university-based hospital for cases of dyskinesia associated with anticonvulsants. Two additional patients, one adult and one pediatric patient were identified. Three patients developed chorea while receiving combination anticonvulsants. Two patients had transient chorea that resolved with withdrawal of one of the drugs. All three patients were using phenytoin and lamotrigine in combination when the chorea started, chorea improved with tapering one of the medications. Polytherapy with certain anticonvulsants may predispose patients to drug-induced chorea. A particular increased risk was seen with combinations that have phenytoin and lamotrigine. This could be due to an additive or a synergistic effect on central dopaminergic pathways.

Expression of the sea urchin MyoD homologue, SUM1, is not restricted to the myogenic lineage during embryogenesis.

SUM1 (sea urchin myogenic factor 1) is a sea urchin homologue of the myogenic basic helix-loop-helix transcription factors of the MyoD family. SUM1 was initially cloned from Lytechinus variegatus where immunocytochemistry demonstrated restricted expression in precursors of the circumesophageal muscles, the only identified muscle cells in the early embryo. Subsequent in situ hybridization analysis indicates that SUM1 embryonic expression is not restricted to the myogenic lineage; a distinct population of nonmyogenic cells also expresses SUM1. For comparative purposes, we cloned the SUM1 orthologue in the distantly related sea urchin, Strongylocentrotus purpuratus, where we found SpSUM1 transcripts in the same population of nonmyogenic cells.

Photoparoxysmal response in late infantile neuronal ceroid-lipofuscinosis.

A patient presented with rapid developmental regression whose MRI findings suggested a leukodystrophy, but nerve, muscle, skin, and bone marrow biopsies were unrevealing. A characteristic photoparoxysmal response on electroencephalogram provided an important clue for the correct diagnosis of late infantile neuronal ceroid-lipofuscinosis, which was confirmed later with electron microscope examination of a brain biopsy. In patients with rapid neurologic deterioration, diagnosis of neuronal ceroid-lipofuscinosis should be considered and an electroencephalogram should be performed using photic stimulation to look for characteristic findings.

Evolution of the chordate muscle actin gene.

The ascidians Styela plicata, S. clava, and Mogula citrina are urochordates. The larvae of urochordates are considered to morphologically resemble the ancestral vertebrate. We asked whether larval and adult ascidian muscle actin sequences are nonmusclelike as in lower invertebrates, musclelike as in vertebrates, or possess characteristics of both. Nonmuscle and muscle actin cDNA clones from S. plicata were sequenced. Based on 27 diagnostic amino acids, which distinguish vertebrate muscle actin from other actins, we found that the deduced protein sequences of ascidian muscle actins exhibit similarities to both invertebrate and vertebrate muscle actins. A comparison to muscle actins from different vertebrate and invertebrate phylogenetic groups suggested that the urochordate muscle actins represent a transition from a nonmusclelike sequence to a vertebrate musclelike sequence. The ascidian adult muscle actin is more similar to skeletal actin and the larval muscle actin is more similar to cardiac actin, which indicates that the divergence of the skeletal and cardiac isoforms occurred before the emergence of urochordates. The muscle actin gene may be a powerful probe for investigating the chordate lineage.

Multiple actin genes encoding the same alpha-muscle isoform are expressed during ascidian development.

Ascidian embryos develop rapidly into tadpole larvae containing striated tail muscle cells. We have isolated and characterized five actin cDNA clones from a Styela clava tailbud-stage library. The nucleotide sequences of these clones and genomic Southern blot analysis indicate that they represent at least four different muscle actin genes, which are designated ScTb1, ScTb24, ScTb30, and ScTb12/34. The derived protein sequences of these genes indicate that they encode the same alpha-muscle actin which has features related to each of the three different classes of vertebrate alpha-muscle actins. Northern and in situ hybridization with probes prepared from the 3' untranslated region (UTR) of several of the ScTb clones showed that these muscle actin genes are expressed in different temporal and spatial patterns during development. ScTb1 was detected in eggs, embryos, and adults, ScTb24 and ScTb12/34 were detected in embryos and adults, and ScTb30 was detected only in embryos. The maternal transcripts disappeared shortly after fertilization and zygotic mRNAs were first detected during gastrulation and continued to accumulate during subsequent tail muscle differentiation. ScTb30 mRNA, which is expressed in the embryo, peaks during the tailbud stage and is present at low levels in the tadpole larva. In contrast, ScTb1, ScTb24, and ScTb12/34 mRNAs, which are expressed in embryos and adults, peak during the late tailbud stage and are present in substantial quantities in the larva. The ScTb24 gene was detected only in tail muscle cells, whereas the ScTb30 gene was detected in embryonic tail muscle, mesenchyme, epidermal, and neural cells. The ScTb24 mRNA also accumulates primarily in vascular tissue in the branchial sac and mantle of adults. The existence of a gene family encoding the same alpha-muscle actin isoform is unique among the chordates and may function to maximize muscle actin production during the rapid differentiation phase of ascidian larval muscle cells.

Temporal and spatial expression of a cytoskeletal actin gene in the ascidian Styela clava.

We have cloned and characterized the temporal and spatial expression of ScCA15, a cDNA clone encoding an actin gene in the ascidian Styela clava. The partial nucleotide and derived amino acid sequences of this singlecopy gene suggest that it is a cytoskeletal actin. Northern analysis shows that ScCA15 corresponds to a 1.8-kb mRNA that is transcribed during oogenesis, during embryonic development, and in the adult. In situ hybridization shows that maternal ScCA15 mRNA is distributed uniformly in the cytoplasm of the oocyte and unfertilized egg. During the period of ooplasmic segregation following fertilization, however, ScCA15 mRNA appears to be translocated into the ectoplasm, a specialized cytoplasmic region of the egg. During the early cleavages, the ectoplasmic transcripts are partitioned to ectodermal cells in the animal hemisphere, which are precursors of the epidermis and nervous system of the larva. Maternal ScCA15 mRNA is degraded just before gastrulation and replaced by zygotic transcripts which begin to accumulate between the neurula and mid-tailbud stages. Zygotic ScCA15 mRNA accumulates primarily in the epidermal and neural cells, although lower levels of these transcripts may also be present in tail muscle cells. These results show that two mechanisms are used to concentrate ScCA15 mRNA in the ectodermal cells during development: 1) localization and differential segregation of maternal transcripts and 2) specific expression of the ScCA15 gene. ScCA15 mRNA is detected by in situ hybridization in the testes, ovaries, alimentary tract, and endostyle of adults. In the testes, ScCA15 mRNA is present in developing sperm, whereas in the ovary, these transcripts are present in the germinal epithelium and developing oocytes. In the alimentary tract, ScCA15 mRNA is confined to the gastric epithelium of the esophagus, stomach, and intestine. Since the ScCA15 gene is expressed in embryonic and adult tissues that are undergoing rapid cell division, this actin is likely to function in some aspect of cell proliferation.

Monoclonal antibody detects embryonic epitope specific for nerve-derived transferrin.

Monoclonal antibodies were generated against transferrin purified from chick embryo extract by fusing spleen cells from BALB/c mice immunized against embryonic transferrin, with myeloma cells. Antibodies produced by the selected hybridoma clones were all type IgG. Twelve clones were selected for secretion of antibodies to the embryo extract-derived transferrin, and three clones were studied extensively. Immunoblotting was used to demonstrate antibody binding to several avian transferrin proteins derived from adult chicken serum, adult chicken peripheral nerves, and ovotransferrin. Screening and detailed epitope analysis were accomplished by solid-phase immunoassay. The results indicated that two clones, 2G9.1 and 2B11.1, recognized the embryonic and egg antigens in preference to the adult proteins. However, a third clone, 6H2.1, recognized the nerve-derived transferrin preferentially to both the embryonic and adult serum antigens. None of the clones recognized the serum-derived transferrin in preference to the other antigens. These results indicate that embryonic epitope(s) are conserved in the nerve- but not the serum-derived transferrin. They also show that the neural antigen has site(s) distinct from the embryonic proteins. No changes in displacement curves were observed after these proteins were digested with neuraminidase, indicating that the epitope differences discovered are not intimately related to sialic acid residues on the various transferrins.

Differential expression of a muscle actin gene in muscle cell lineages of ascidian embryos.

Specific probes were used to examine the accumulation of muscle actin mRNA during embryonic development of the ascidian Styela. Clones of a muscle actin gene were obtained from an adult mantle cDNA library. Four lines of evidence indicate that these clones correspond to a muscle actin gene. First, their coding regions share 11 of 14 diagnostic amino acid positions with mammalian smooth and skeletal muscle actins. Second, subclones that contain only the 3' noncoding region of the gene select mRNA coding for muscle actin, while subclones that include the coding region of the gene select mRNA coding for muscle and nonmuscle actins. Third, a probe that contains only the 3' noncoding region detects a single band, corresponding to a 2 kb transcript, while a probe that includes the coding region detects the 2 kb transcript and at least one other band, presumably a cytoplasmic actin transcript. Fourth, the 3' noncoding region probe detects transcripts only in muscle cells and their precursors, while the coding region probe detects transcripts in muscle and nonmuscle cells. The muscle actin transcript is present at very low levels in eggs and early embryos, begins to accumulate between the early gastrula and tailbud stages, and by the tadpole stage attains a level about 25-fold higher than in the egg. In situ hybridization showed that embryonic muscle actin transcripts are restricted to the muscle cell lineages. These transcripts were initially observed in primary muscle lineage cells (descendants of the B4.1 blastomeres) at the early gastrula stage and continued to be present in these cells throughout embryonic development. In contrast, muscle actin transcripts did not appear in secondary muscle lineage cells (descendants of b4.2 and A4.1 blastomeres) until the mid-tailbud stage, and were not detected in mesenchyme cells, the presumptive adult muscle cell precursors, at any time during embryonic development. The results suggest that muscle actin gene expression is subject to spatial and temporal regulation in the muscle cell lineages.

Collagenase activity in skin fibroblasts of patients with amyotrophic lateral sclerosis.

In the current study we have measured collagenase activity released from skin explants and fibroblasts of patients with both Guam-type and sporadic amyotrophic lateral sclerosis and controls. The rationale for such a study derives from work reported more than 20 years ago demonstrating abnormalities in skin collagen metabolism in patients with the disease. We were not able to find significant differences in collagenase activity when fibroblasts were compared relative to the total protein secreted. This is explained, in part, by our finding of an increase in total protein released from fibroblasts of the amyotrophic lateral sclerosis patient group. Increased collagenase release did occur when activity was expressed per number of cells plated but was not statistically significant. In addition, increased release followed a 3-day lag period in skin organ culture. These results suggest that collagenase and other enzymes known to activate collagenase, such as plasminogen activator, capable of degrading extracellular matrix components might be responsible for the increased collagenolytic activity previously observed in amyotrophic lateral sclerosis patients' skin. Further evaluation of extracellular-acting degradative enzymes from skin, muscle, nerve and central nervous system may be important to follow-up such leads in understanding the pathogenesis of this enigmatic and fatal disorder.

Extracellular matrix (ECM) synthesis in muscle cell cultures: quantitative and qualitative studies during myogenesis.

When the synthesis of extracellular matrix components was examined in G8-1 murine skeletal muscle cells as a function of differentiation, non-collagen and to an even greater extent collagen synthesis was increased. Specifically, collagen types I, III, IV, laminin and fibronectin were identified by SDS-PAGE. Immunoprecipitation, with specific antibodies revealed that both the cell layer and medium of differentiated multinucleated myotubes contained increased levels of type IV collagen and laminin, decreased levels of type III collagen and fibronectin and equivalent levels of type I collagen compared to mononuclear myoblasts.

Specificity of chicken and mammalian transferrins in myogenesis.

Chicken transferrins isolated from eggs, embryo extract, serum or ischiatic-peroneal nerves are able to stimulate incorporation of [3H]thymidine, and promote myogenesis by primary chicken muscle cells in vitro. Mammalian transferrins (bovine, rat, mouse, horse, rabbit, and human) do not promote [3H]thymidine incorporation or myotube development. Comparison of the peptide fragments obtained after chemical or limited proteolytic cleavage demonstrates that the four chicken transferrins are all indistinguishable, but they differ considerably from the mammalian transferrins. The structural differences between chicken and mammalian transferrins probably account for the inability of mammalian transferrins to act as mitogens for, and to support myogenesis of, primary chicken muscle cells.

Extracellular-matrix synthesis by skeletal muscle in culture. Major secreted collagenous proteins of clonal myoblasts.

We have previously shown that G8-1, a murine clonal skeletal-muscle cell line, produces a substrate-attached extracellular matrix [Beach, Burton, Hendricks & Festoff (1982) J. Biol. Chem. 257, 11437-11442]. To examine further the expression of extracellular-matrix proteins by muscle cells, we have analysed the collagenous proteins secreted by G8-1 myoblasts. We have found that collagens and/or procollagens, corresponding to genetic types I, III and IV (and possibly V), are produced and secreted by G8-1 myoblasts. The major secreted collagenous polypeptides were identified as alpha 1 type I and its precursors by using pulse-chase studies, pepsin and collagenase digestions and CNBr fragmentation. The presence of lesser amounts of the other collagens was determined by immunoprecipitation. These results demonstrate that clonal skeletal-muscle cells, in the absence of fibroblasts and an exogenous collagen substrate, are able to synthesize and secrete several extracellular-matrix collagenous proteins in proportions similar to those which are commonly found in muscle tissue and mixed cultures of muscle cells and fibroblasts.

Peripheral nerve extract promotes long-term survival and neurite outgrowth in cultured spinal cord neurons.

The hypothesis that peripheral, skeletal muscle tissue contains a trophic factor supporting central neurons has recently been investigated in vitro by supplementing the culture medium of spinal cord neurons with muscle extracts and fractions of extract. We extended these studies asking whether or not a trophic factor is present in peripheral nerves, the connecting link between muscle and central neurons via which factors may be translocated from muscle to neurons by the retrograde transport system. Lumbar, 8-day-old chick spinal cords were dissociated into single cells and then cultured in the presence of peripheral nerve extract. Cytosine arabinoside was added to inhibit proliferation of nonneuronal cells. In the presence of nerve extract, spinal cord neurons survived for more than a month, extended numerous neurites, and showed activity of choline acetyltransferase. In the absence of extract, neurons attached and survived for a few days but then died subsequently in less than 10 days. Neurite outgrowth did not occur in the absence of extract. Withdrawal of extract from the medium of established neuronal cultures caused progressive loss of both cells and neurites. Other tissues also contained neuron supporting activity but less than that found in nerve extract. These studies indicate that peripheral nerves contain relatively high levels of spinal cord neuron-directed trophic activity, suggesting translocation of neurotrophic factor from muscle to central target neurons. The neurotrophic factor has long-term (weeks) effects, whereas short-term (days) survival is factor independent.

Appearance of acetylcholinesterase molecular forms in noninnervated cultured primary chick muscle cells.

Asymmetric forms of AChE have generally not been detected in cultured chick skeletal muscle cells in the absence of cocultured neurons. To explore further neurotrophic effects of adult peripheral nerve extracts (NE) on muscle in vitro, we reexamined the appearance of various molecular forms of AChE in cultured chick muscle cells in the presence of NE. The various molecular forms of AChE were distinguished by sucrose gradient sedimentation and radioenzymatic techniques. In the presence of NE, cells proliferated during the first 48 hr of culture, then fused and formed spontaneously contracting myotubes by 6-8 days in culture. Total AChE, 5.4 S, and 11.5 S molecular forms reached activity plateaus by 8 days in culture which persisted until cultures were terminated at day 20. Between 1 and 6 days in culture, 19.5 S AChE (A12) was not detected. The A12 form was first observed at 7 days reaching a maximum of 11.3% of the total AChE at 14 days and then gradually declined to a level of 1% at day 20. Since the A12 form declined in older cultures but comprised 25% in embryonic muscle tissue, we examined the possible requirement of neurons in culture to attain higher levels of A12 AChE. Spinal cord neurons were plated onto 6-day muscle cultures and AChE activities were measured between 8 and 20 days. The results showed that 19.5 S AChE activity in the presence of both spinal cord neurons and NE was no greater than that found in the presence of NE alone. To suppress spontaneous contraction, 0.6 microM tetrodotoxin (TTX) or 15 microM d-tubocurarine (dTC) were added to 5-day-old muscle cultures at a time when myotubes were differentiated but contractile activity had not begun. TTX had cytotoxic effects and inhibited further development of myotubes. In contrast, dTC had no deleterious effect on morphological development, eliminated contraction, but did not interfere with the appearance of any forms of AChE including the A12 form. These studies show that primary chick muscle cells are capable of producing the A12 form of AChE if cultured in NE-supplemented medium. In this culture system, production of the A12 form does not require activity or innervation.

Acetylcholine receptor turnover in clonal muscle cells: role of plasmin and effects of protease inhibitors.

Characteristics of acetylcholine receptors were evaluated in G8-1, a continuous skeletal muscle line. Peak binding of 125I-alpha-bungarotoxin was in 10-day-old contractile myotubes at 4-8 nm. Turnover was studied using two different methods; both indicated half-times as little as half as long as previously reported for primary cultures. The effects of a variety of protease inhibitors on receptor turnover were assessed to determine if G8-1 receptors were less stable or turned over faster because of increased neutral protease activity. Leupeptin, antipain, and chloroquine markedly slowed receptor degradation. Inhibitors of plasmin or plasminogen activator had definite but less dramatic effects on receptor turnover. Results from studies in which plasmin was increased in the tissue culture media indicated that a small but definite acceleration of receptor turnover occurred. In clonal G8-1 cells, total number of acetylcholine receptors is controlled by negative feedback and although the major pathway for receptor degradation is lysosomal, plasmin may play a role in initiating receptor internalization.

The identification of neurotrophic factor as a transferrin.

Partially purified neurotrophic factor (NTF) from chicken nerves comigrated with transferrin and a component in several preparations known to have neurotrophic effects on cultured skeletal muscle cells. One-dimensional gel electrophoretograms of proteolytic fragments of NTF and fragments obtained from transferrins purified from chicken eggs, serum and embryos were indistinguishable. These purified transferrins, like NTF, all stimulated the incorporation of [3H]thymidine and supported myotube formation to a similar degree as NTF. These studies suggest that NTF is a transferrin-like protein and that both transferrins and NTF act by initially promoting myoblast proliferation and subsequently supporting myogenesis in chick muscle cultures.

Extracellular matrix synthesis by skeletal muscle in culture. Proteins and effect of enzyme degradation.

Extracellular matrix proteins produced by a mouse skeletal muscle cell line, G8-1, were isolated and characterized. Cultures were incubated with [35S]methionine or [3H]glycine and [3H]proline, and the labeled, substrate-attached proteins were obtained after cellular proteins were extracted by deoxycholate in neutral salt. The labeled matrix was analyzed by gel electrophoresis and fluorography before and after enzymatic digestion. Of the nine major bands present in the matrix, four were identified. Fibronectin and collagen were detected on the bases of their relative mobilities, differential labeling with 3H-versus 35S-labeled amino acids and their solubilization by protease free collagenase. High molecular weight material which was present in the matrix was also sensitive to collagenase and probably included cross-linked collagen and laminin. Proteins co-migrating with actin and myosin were also present in the extracellular matrix. These results are novel in that they demonstrate that the skeletal muscle phenotype, not contaminated with fibroblastic elements, is able to synthesize basal lamina-type macromolecules and incorporate them into an insoluble, extracellular matrix. Since this cell line is able to form functional synaptic contacts with neuronal cells, the influence of nerve on basal lamina production by muscle in vitro is possible.

Identification of cell types in rat hippocampal slices maintained in organotypic cultures.

We have cultured transverse slices of the hippocampal formation from neonatal rats and have identified the cell types which appear in the outgrowth with cell type specific markers. Tetanus toxin and anti-tetanus toxoid, as well as antisera to neurofilaments and 14-3-2 protein, were used to identify neurons. Astrocytes were identified with antisera to glial fibrillary acidic protein and were the predominant non-neural cell type. Fibroblastic cells were labeled with antisera to fibronectin and to myosin and oligodendroglia were identified with antisera to galactocerebroside. The hippocampal neurons could be classified as 1 or the 3 types present in vivo (pyramidal cells, granule cells, or GABAergic interneurons) on the basis of their size, shape, location, or reaction with antisera to glutamic acid decarboxylase. Outgrowth of glial cells and neurites occurred within hours of explantation. Within a few days granule cell neurons migrated onto the glial cell layer from the explant. Their movement is probably related to their migration during in vivo development of the granule cell layer. Synapse formation was observed by electron microscopic analysis beginning about 3-5 days in vitro and areas of neuropil containing many synapses were observed after 3-4 weeks. This culture system should be useful for further studies on the cellular processes which occur during hippocampal development and plasticity.

Identification of myosin in isolated synaptic junctions.

A monospecific antibody prepared against chicken gizzard myosin reacted with only one peptide corresponding to myosin heavy chain (Mr = 200,000) in gels of synaptic plasma membranes (SPM) and synaptic junctions (SJ) prepared from several species. Preadsorption of antisera with purified brain myosin eliminated antibody reactivity to SPMs and SJs. SJs were found to contain approximately 3 times the concentration of myosin found in SPMs when assayed by an indirect immunoradiometric assay. Postsynaptic density and myelin fractions contained no myosin detectable by immunoradiometric assay, antibody binding to gels, or Coomassie blue staining. The band identified as myosin in SJ fraction yielded peptide fingerprints indistinguishable from fingerprints of purified brain myosin but distinct from fingerprints of purified smooth and skeletal muscle myosins. The distribution of exogenous [125I]myosin during subcellular fractionation indicated that myosin in isolated synaptic junction could not have resulted from artifactual re-distribution of soluble myosin. Together these results show that a non-muscle myosin is an endogenous component of CNS asymmetric synapses.