Human Wnt-13 is developmentally regulated during the differentiation of NTERA-2 pluripotent human embryonal carcinoma cells.

The Wnt gene family encodes a series of conserved glycoproteins that regulate pattern formation during embryogenesis, in a variety of tissues including the nervous system. As with other genes that control embryonic cell differentiation, members of the Wnt family have also been implicated in tumourigenesis. To search for Wnt genes involved in human teratocarcinomas, with a possible role in human embryogenesis, we used RT-PCR primed with degenerate oligonucleotides to analyse mRNA from differentiating cultures of the pluripotent human embryonal carcinoma (EC) cell line NTERA-2. NTERA-2 EC cells differentiate into neurons and other cell types when induced with retinoic acid. Wnt gene expression was not detected in the undifferentiated EC cells, but Wnt-related PCR fragments were amplified from differentiating cultures, 4-14 days after induction with retinoic acid. The RT-PCR products were composed primarily of DNA fragments corresponding to the recently identified human Wnt-13 gene. No other Wnt-related genes were identified. Northern analysis confirmed induction of Wnt-13 as a 2.4 kb mRNA during the early phases of retinoic acid-induced differentiation, and during differentiation along a non-neural pathway induced by hexamethylene bisacetamide (HMBA), but not in the terminally differentiated neurons. Wnt-13 remained expressed in non-neural differentiated NTERA-2 cells, even several weeks after the induction of differentiation. The time course of induction, its induction by HMBA, and its persistence in differentiated cells indicate that Wnt-13 expression is not dependent upon direct activation by retinoic acid. Wnt-13 was not detected, or only detected at low levels, in other human EC cells. However, it was found to be expressed at a high level in one malignant teratoma cell line, 577MF, that does not exhibit an EC phenotype although it was derived from a testicular teratocarcinoma. At least two members of the human frizzled gene family, thought to encode receptors for Wnt proteins, were also expressed in the NTERA-2 cells, suggesting the presence of a mechanism by which endogenously expressed Wnt-13 could modulate the histogenesis of teratocarcinomas by mediating interactions between sub-populations of differentiating EC cells. We note that Wnt-13 maps to chromosome 1p13, a region reported to be subject to relatively frequent loss of heterozygosity in germ cell tumours. Further analysis indicated that 465 bp of the published Wnt-13 sequence, within the predicted 5' UTR, is incorrect and is possibly derived from a human mitochondrial DNA sequence.

Analysis of a panel of antibodies to APC reveals consistent activity towards an unidentified protein.

Acquisition of truncating mutations in the adenomatous polyposis coli (APC) protein underlies the progression of the majority of sporadic and familial colorectal cancers. As such, the localisation patterns and interacting partners of APC have been extensively studied in a range of systems, relying on the use of a broad panel of antibodies. Until recently, antibodies to APC have been used largely unchecked. However, several recent reports have been invaluable in clarifying the use of a number of antibodies commonly used to detect APC. Here, we analyse the specificity of a further subset of antibodies to APC. We used a panel of six commercially available antibodies (directed to the amino and carboxy termini of APC) and confirm the detection of full-length APC by immunoblotting. We demonstrate that a 150 kDa protein, also reproducibly detected by this panel of antibodies, is unlikely to be APC. We present data for the immunological staining patterns of the APC antibodies and validate the results through RNAi. Using this approach, we confirm that the apical staining pattern, observed by immunofluorescence and previously reported in cell systems, is unlikely to be APC. Finally, we present our data as a summary of APC-antibody specificities for APC.

Pseudouridine distribution in mammalian 18 S ribosomal RNA. A major cluster in the central region of the molecule.

Human and rodent 18 S rRNA contain about 38 pseudouridine residues. By correlating RNA oligonucleotide data with complete sequence data derived from ribosomal DNA, 30 pseudouridine residues can be located in the RNA sequence, either exactly or to within two or three residues. Pseudouridine and 2'-O-methyl groups are interspersed throughout mammalian 18 S rRNA, but not in closely parallel fashion. Whereas the largest cluster of 2'-O-methyl groups is in the 5' one-third of the molecule, the greatest concentration of pseudouridine is in the central one-third of the molecule.

28 S ribosomal RNA in vertebrates. Locations of large-scale features revealed by electron microscopy in relation to other features of the sequences.

The 28 S rRNA from several vertebrate species, when examined by electron microscopy, is seen to contain regions of extensive secondary structure, as first reported for HeLa-cell 28 S rRNA by Wellauer & Dawid [(1973) Proc. Natl. Acad. Sci. U.S.A. 70, 2827-2831]. Here we correlate the locations of these regions, determined from the electron-microscopic data, with the primary structure of 28 S rRNA from human, mouse and Xenopus laevis determined by sequence analysis of rDNA. The secondary-structure features observed by electron microscopy correspond closely to phylogenetically variable G + C-rich regions that largely comprise the eukaryotic expansion segments in these three species. In most if not all cases the features can be identified with long G + C-rich helices deduced from sequence data. Correlations are given between the locations of the secondary-structure features and several 'landmark' restriction sites in 28 S rDNA. By correlating the locations of the rRNA methyl groups reported elsewhere [Maden (1988) J. Mol. Biol. 201, 289-314] with the present findings it is concluded that the rRNA secondary-structure features revealed by electron microscopy are largely or wholly unmethylated.

Interaction between Ku80 protein and a widely used antibody to adenomatous polyposis coli.

The adenomatous polyposis coli (APC) gene and its expressed product are highly studied because of its role as a tumour-suppressor protein. Inherited mutations in APC lead to the condition known as familial adenomatous polyposis (FAP), which predisposes the affected individuals to colorectal cancer. Furthermore, mutations in APC are found in the majority of sporadic cases of colon cancer. There have been many published studies concerning the cellular localisation of APC, this being fundamental to our understanding of its function, but there has also been much concern over the specificity of certain commercially available antibodies to APC. Here we report that the widely used antibody APC(N15) demonstrates a strong interaction with the Ku80 subunit of the Ku heterodimer under defined experimental conditions. Based on the data presented here, we suggest that APC(N15) is not suitable for many applications used for the study of APC.

MAL mRNA is induced during the differentiation of human embryonal carcinoma cells into neurons and is also localised within specific regions of the human brain.

We have found that the MAL gene, which encodes a membrane proteolipid expressed during the late stages of T-lymphocyte maturation, is also activated during neuronal differentiation of NTERA2 human embryonal carcinoma cells following induction with retinoic acid. An RT-PCR fragment with a sequence identical to MAL was found amongst 30 cloned DNA fragments corresponding to genes putatively activated during NTERA2 differentiation and isolated using a differential display PCR screen. PCR and Northern blot analysis with a cloned MAL cDNA as a probe confirmed that MAL is not expressed by undifferentiated NTERA2 EC cells, but is expressed, predominantly as a 1.1 kb transcript, within 7 days of retinoic acid-induced differentiation and later in the post-mitotic neurons arising in such cultures. MAL was not expressed in the non-neuronal lineages induced by treatment of NTERA2 cells with the gratuitous inducer hexamethylene bisacetamide. Analysis of cDNA libraries constructed from EC cells, purified neurons and a sub-population of non-neuronal cells (ME311+), confirmed that expression of the MAL gene is activated in the neural lineage of NTERA2 differentiation. Using in situ hybridisation we found that MAL is expressed in the human CNS and especially in grey matter of the cerebral cortex, with less in the cerebellum and the amygdala and little or none in subcortical white matter. In contrast to reports concerning the expression pattern of a rat MAL homologue, MAL was expressed in the human brain predominantly in cell bodies which include neurons, correlating with in vitro data from the NTERA2 line.

Clones of human ribosomal DNA containing the complete 18 S-rRNA and 28 S-rRNA genes. Characterization, a detailed map of the human ribosomal transcription unit and diversity among clones.

We have isolated several new clones of human ribosomal DNA. Each clone contains part of the external transcribed spacer, a complete 18 S-rRNA gene, the internal transcribed spacers, a complete 28 S-rRNA gene and a short downstream flanking region. We present a detailed map of the human ribosomal transcription unit with the locations of numerous useful restriction sites. In particular, a unique NheI site in the 5.8 S-rRNA gene enabled this gene to be mapped with respect to the 18 S-rRNA and 28 S-rRNA genes. The human 45 S-rRNA coding region is approx. 13,000 nucleotide residues long, of which the external transcribed spacer comprises approx. 3700 nucleotide residues and the first and second internal transcribed spacers comprise approx. 1070 and 1200 nucleotide residues respectively. A partial survey for sites of variation between clones has revealed a single point of variation among 18 S-rRNA gene sequences (a T/C variation at position 140), several sites of length variation in the regions of the transcribed spacers closely flanking the 18 S-rRNA genes, and some sites of length variation among 28 S-rRNA genes. Most of these sites of variation are associated with simple sequence tracts and are in regions that are known to undergo relatively rapid evolutionary divergence. In particular, the sites of variation among 28 S-rRNA genes occur in G + C-rich tracts whose lengths vary among vertebrates and that can be correlated with extensive hairpin structures previously observed by electron microscopy. Each of the clones so far surveyed in detail differs from the others in one or more respects.

Regulation of intracellular Ca2+ in response to muscarinic and glutamate receptor agonists during the differentiation of NTERA2 human embryonal carcinoma cells into neurons.

Single cell microfluorimetry was used to study intracellular calcium ion signals ([Ca(2+)](i)) evoked by acetylcholine (ACh), glutamate receptor agonists and by KCI-induced membrane depolarization, during neuronal differentiation of the human embryonal carcinoma (EC) cell line, NTERA2. In undifferentiated NTERA2 EC cells, [Ca(2+)](i) was elevated in response to ACh, but not to the glutamate receptor agonists NMDA, kainate or AMPA. The ACh-induced rise in [Ca(2+)](i) was dependent upon both Ca(2+) influx and Ca(2+) mobilization from cytoplasmic calcium stores. Three other human EC cell lines responded similarly to ACh but not to glutamate or KCI-induced depolarization. In neurons derived from NTERA2 cells by retinoic acid induction, [Ca(2+)](i) signals were evoked by ACh, NMDA, kainate and by an elevation of the extracellular KCI concentration. As in undifferentiated EC cells, the ACh-mediated increases in [Ca(2+)](i) were governed by both Ca(2+) influx and Ca(2+) mobilization. In contrast, the effects of NMDA, kainate and KCI did not involve intracellular Ca(2+) mobilization. The appearance of glutamate and KCI responsiveness was not detected in non-neuronal differentiated derivatives of NTERA2 cells. Using a number of pharmacologically defined muscarinic receptor antagonists we found that NTERA2 EC cells express M(1), M(3), M(4) and possibly M(5) receptor subtypes linked to changes in [Ca(2+)](i), whilst only M(3) and M(5) are present in NTERA2-derived neurons. The results were supported by PCR analysis of the muscarinic mRNA species expressed in the cells. The data demonstrate that differentiation of NTERA2 EC cells into neurons involves the induction of functional glutamate receptors coupled to rises in [Ca(2+)](i), and changes in the expression of muscarinic ACh receptor subtypes.