Multivariate Analysis of RNA Chemistry Marks Uncovers Epitranscriptomics-Based Biomarker Signature for Adult Diffuse Glioma Diagnostics

One of the main challenges in cancer management relates to the discovery of reliable biomarkers, which could guide decision-making and predict treatment outcome. In particular, the rise and democratization of high-throughput molecular profiling technologies bolstered the discovery of “biomarker signatures” that could maximize the prediction performance. Such an approach was largely employed from diverse OMICs data (i.e., genomics, transcriptomics, proteomics, metabolomics) but not from epitranscriptomics, which encompasses more than 100 biochemical modifications driving the post-transcriptional fate of RNA: stability, splicing, storage, and translation. We and others have studied chemical marks in isolation and associated them with cancer evolution, adaptation, as well as the response to conventional therapy. In this study, we have designed a unique pipeline combining multiplex analysis of the epitranscriptomic landscape by high-performance liquid chromatography coupled to tandem mass spectrometry with statistical multivariate analysis and machine learning approaches in order to identify biomarker signatures that could guide precision medicine and improve disease diagnosis. We applied this approach to analyze a cohort of adult diffuse glioma patients and demonstrate the existence of an “epitranscriptomics-based signature” that permits glioma grades to be discriminated and predicted with unmet accuracy. This study demonstrates that epitranscriptomics (co)evolves along cancer progression and opens new prospects in the field of omics molecular profiling and personalized medicine.

Cellular and molecular characterization of IDH1-mutated diffuse low grade gliomas reveals tumor heterogeneity and absence of EGFR/PDGFRα activation

Diffuse low grade gliomas (DLGG, grade II gliomas) are slowly-growing brain tumors that often progress into high grade gliomas. Most tumors have a missense mutation for IDH1 combined with 1p19q codeletion in oligodendrogliomas or ATRX/TP53 mutations in astrocytomas. The phenotype of tumoral cells, their environment and the pathways activated in these tumors are still ill-defined and are mainly based on genomics and transcriptomics analysis. Here we used freshly-resected tumors to accurately characterize the tumoral cell population and their environment. In oligodendrogliomas, cells express the transcription factors MYT1, Nkx2.2, Olig1, Olig2, Sox8, four receptors (EGFR, PDGFRα, LIFR, PTPRZ1) but not the co-receptor NG2 known to be expressed by oligodendrocyte progenitor cells. A variable fraction of cells also express the more mature oligodendrocytic markers NOGO-A and MAG. DLGG cells are also stained for the young-neuron marker doublecortin (Dcx) which is also observed in oligodendrocytic cells in nontumoral human brain. In astrocytomas, MYT1, PDGFRα, PTPRZ1 were less expressed whereas Sox9 was prominent over Sox8. The phenotype of DLGG cells is overall maintained in culture. Phospho-array screening showed the absence of EGFR and PDGFRα phosphorylation in DLGG but revealed the strong activation of p44/42 MAPK/ERK which was present in a fraction of tumoral cells but also in nontumoral cells. These results provide evidence for the existence of close relationships between the cellular phenotype and the mutations found in DLGG. The slow proliferation of these tumors may be associated with the absence of activation of PDGFRα/EGFR receptors.

Anatomical study of serotonergic innervation and 5-HT(1A) receptor in the human spinal cord

Serotonergic innervation of the spinal cord in mammals has multiple roles in the control of motor, sensory and visceral functions. In rats, functional consequences of spinal cord injury at thoracic level can be improved by a substitutive transplantation of serotonin (5-HT) neurons or regeneration under the trophic influence of grafted stem cells. Translation to either pharmacological and/or cellular therapies in humans requires the mapping of the spinal cord 5-HT innervation and its receptors to determine their involvement in specific functions. Here, we have performed a preliminary mapping of serotonergic processes and serotonin-lA (5-HT_1A) receptors in thoracic and lumbar segments of the human spinal cord. As in rodents and non-human primates, 5-HT profiles in human spinal cord are present in the ventral horn, surrounding motoneurons, and also contact their presumptive dendrites at lumbar level. 5-HT_1A receptors are present in the same area, but are more densely expressed at lumbar level. 5-HT profiles are also present in the intermediolateral region, where 5-HT_1A receptors are absent. Finally, we observed numerous serotonergic profiles in the superficial part (equivalent of Rexed lamina II) of the dorsal horn, which also displayed high levels of 5-HT_1A receptors. These findings pave the way for local specific therapies involving cellular and/or pharmacological tools targeting the serotonergic system.

The human NTERA2 neural cell line generates neurons on growth under neural stem cell conditions and exhibits characteristics of radial glial cells

NTERA2 cells are a human neural cell line generating neurons after exposure to retinoic acid and, as such, are widely used as a model of neurogenesis. We report that these cells form spheres when grown in serum-free medium supplemented with basic fibroblast growth factor (bFGF) and epidermal growth factor (EGF). These spheres were found to express markers of radial glial cells such as, Pax6, glutamate transporter (GLAST), tenascin C, brain lipid-binding protein (BLBP), and the 3CB2 antigen. On plating on an adhesive substrate, NTERA2 spheres generate a large percentage of immature neurons (30-50%) together with a minority of cells of the oligodendrocyte lineage. Thus NTERA2 cells share properties with neural stem cells. However, at variance with the latter, we found that they produce their own bFGF implicated in an autocrine or paracrine proliferative loop and that they do not generate astrocytes after differentiation. These results provide an interesting model to study radial glial cells and their role in human neurogenesis.

Direct cell-cell interactions control apoptosis and oligodendrocyte marker expression of neuroepithelial cells

During brain development, the neuroepithelium generates neurons and glial cells. Proliferation and differentiation of neuroepithelial cells are controlled by a complex combination of secreted factors and more intrinsic or local mechanisms, such as lateral inhibition and asymmetric division. To obtain further insights into the signals governing neuroepithelial cell fate, we used the immortomouse to derive conditionally immortalised cell lines from mouse E10 neuroepithelium. We isolated a nestin-positive basic fibroblast growth factor (bFGF)-responsive cell line (SVE10-23) which mostly differentiate into astrocytes when cocultured with primary cortical cells. We found that, by simply lowering the cell density, SVE10-23 cells embarked on oligodendrocytic differentiation as indicated by the strong expression of galactocerebroside C and 2'3'-cyclic nucleotide 3'-phosphodiesterase. Apoptosis accompanied the differentiation, and all cells died within 1 week. We present here evidence that direct interactions between cells are the main mechanism regulating this oligodendrocytic differentiation. We demonstrate that SVE10-23 cells contact or proximity inhibit their differentiation, prevent apoptosis, and promote their proliferation. Similarly, others nestin-positive precursor cell lines and nonimmortalised bFGF-grown E10 cells were found to spontaneously differentiate at low density, thus generalising the idea that neural precursor fate is regulated by direct cell-cell interactions. The SVE10-23 cell line provides a valuable tool with which to study further the molecular components implicated in this mode of regulation.

Regulation of glial differentiation of MHP36 neural multipotent cell line

MHP36 is a nestin bFGF-dependent cell line isolated from embryonic hippocampus using a thermolabile form of SV40 T antigen. When grafted in ischemic hippocampus MHP36 cells differentiate and alleviate the cognitive deficit associated with the lesion. We report here in vitro features of MHP36 cells. First, we found that T Ag expression was not necessary for MHP36 growth as cells cultured at the nonpermissive temperature carry on proliferating at a normal rate, Second, we observed that part of MHP36 cells spontaneously differentiate into astrocytes when bFGF is removed at39 degrees C. This differentiation was increased 4-fold by leukemia inhibitory factor. Third, we found that the majority of cells spontaneously expressed oligodendrocytic markers (CNPase, A2B5, GalC) when cultured at low density.

New modulatory alpha subunits for mammalian Shab K+ channels

Two novel K+ channel alpha subunits, named Kv9.1 and Kv9.2, have been cloned. The Kv9.2 gene is situated in the 8q22 region of the chromosome. mRNAs for these two subunits are highly and selectively expressed in the nervous system. High levels of expressions are found in the olfactory bulb, cerebral cortex, hippocampal formation, habenula, basolateral amygdaloid nuclei, and cerebellum. Interestingly Kv9.1 and Kv9.2 colocalized with Kv2.1 and/or Kv2.2 alpha subunits in several regions of the brain. Neither Kv9.1 nor Kv9.2 have K+ channel activity by themselves, but both modulate the activity of Kv2.1 and Kv2.2 channels by changing kinetics and levels of expression and by shifting the half-inactivation potential to more polarized values. This report also analyzes the changes in electrophysiological properties of Kv2 subunits induced by Kv5.1 and Kv6.1, two other modulatory subunits. Each modulatory subunit has its own specific properties of regulation of the functional Kv2 subunits, and they can lead to extensive inhibitions, to large changes in kinetics, and/or to large shifts in the voltage dependencies of the inactivation process. The increasing number of modulatory subunits for Kv2.1 and Kv2.2 provides an amazingly new capacity of functional diversity.

Modes of regulation of shab K+ channel activity by the Kv8.1 subunit

The Kv8.1 subunit is unable to generate K+ channel activity in Xenopus oocytes or in COSm6 cells. The Kv8.1 subunit expressed at high levels acts as a specific suppressor of the activity of Kv2 and Kv3 channels in Xenopus oocytes (Hugnot, J. P., Salinas, M., Lesage, F., Guillemare, E., Weille, J., Heurteaux, C., Mattéi, M. G., and Lazdunski, M. (1996) EMBO J. 15, 3322-3331). At lower levels, Kv8.1 associates with Kv2.1 and Kv2.2 to form hybrid Kv8.1/Kv2 channels, which have new biophysical properties and more particularly modified properties of the inactivation process as compared with homopolymers of Kv2.1 or Kv2.2 channels. The same effects have been seen by coexpressing the Kv8.1 subunit and the Kv2.2 subunit in COSm6 cells. In these cells, Kv8.1 expressed alone remains in intracellular compartments, but it can reach the plasma membrane when it associates with Kv2.2, and it then also forms new types of Kv8.1/Kv2. 2 channels. Present results indicate that Kv8.1 when expressed at low concentrations acts as a modifier of Kv2.1 and Kv2.2 activity, while when expressed at high concentrations in oocytes it completely abolishes Kv2.1, Kv2.2, or Kv3.4 K+ channel activity. The S6 segment of Kv8.1 is atypical and contains the structural elements that modify inactivation of Kv2 channels.

The human inward rectifying K+ channel Kir 2.2 (KCNJ12) gene: gene structure, assignment to chromosome 17p11.1, and identification of a simple tandem repeat polymorphism

K+ channels are essential for a variety of cellular functions in both excitable and nonexcitable cells, and K+ channel gene alteration has been recently described in cardiac and neurological disorders. To explore further the relations between hereditary human diseases and K+ channels, we isolated from a human cosmid library the gene encoding the inwardly rectifying K+ channel alpha-subunit Kir 2.2 (KCNJ12). PCR analysis performed on this clone indicates that the entire open reading frame is contained in one unique exon. A polymorphic (CA)16 sequence was localized 2.2 kb upstream of the ATG start codon. Fluorescence in situ hybridization on human metaphases assigns the gene to band 17p11.1. The implication of a deletion of the Kir 2.2 gene in the Smith-Magenis syndrome, which is also localized at 17p11, is unlikely since a Kir 2.2-linked microsatellite sequence could be amplified from the DNA of a Smith-Magenis syndrome affected patient bearing a 17p interstitial deletion.

Kv8.1, a new neuronal potassium channel subunit with specific inhibitory properties towards Shab and Shaw channels

Outward rectifier K+ channels have a characteristic structure with six transmembrane segments and one pore region. A new member of this family of transmembrane proteins has been cloned and called Kv8.1. Kv8.1 is essentially present in the brain where it is located mainly in layers II, IV and VI of the cerebral cortex, in hippocampus, in CA1-CA4 pyramidal cell layer as well in granule cells of the dentate gyrus, in the granule cell layer and in the Purkinje cell layer of the cerebellum. The Kv8.1 gene is in the 8q22.3-8q24.1 region of the human genome. Although Kv8.1 has the hallmarks of functional subunits of outward rectifier K+ channels, injection of its cRNA in Xenopus oocytes does not produce K+ currents. However Kv8.1 abolishes the functional expression of members of the Kv2 and Kv3 subfamilies, suggesting that the functional role of Kv8.1 might be to inhibit the function of a particular class of outward rectifier K+ channel types. Immunoprecipitation studies have demonstrated that inhibition occurs by formation of heteropolymeric channels, and results obtained with Kv8.1 chimeras have indicated that association of Kv8.1 with other types of subunits is via its N-terminal domain.

Heterologous multimeric assembly is essential for K+ channel activity of neuronal and cardiac G-protein-activated inward rectifiers

The family of G-protein-activated inward-rectifiers K+ channels presently comprise at least 3 cloned members called GIRK1, GIRK2 and GIRK3. A close structural parent of GIRK channels has recently been described as being an ATP-sensitive K+ channel. This paper shows that Xenopus expression of this new channel that we call GIRK4 does not produce an ATP-inhibitable activity with a pharmacological activation by pinacidil as previously described but instead a G-protein activated inward-rectifier. While oocyte expression of single subunits is infrequent and relatively modest in intensity, expression of GIRK1,2, GIRK1,4 and GIRK2,4 mixtures leads to routine and robust expression of K+ channels indicating that heterologous subunit assembly is necessary for activity. Activity of GIRK1,2, GIRK1,4 and GIRK2,4 channels required the presence of ATP acting as an activator at the cytoplasmic face and is further activated by the beta gamma subunits.

Cloning provides evidence for a family of inward rectifier and G-protein coupled K+ channels in the brain

MbIRK3, mbGIRK2 and mbGIRK3 K+ channels cDNAs have been cloned from adult mouse brain. These cDNAs encode polypeptides of 445, 414 and 376 amino acids, respectively, which display the hallmarks of inward rectifier K+ channels, i.e. two hydrophobic membrane-spanning domains M1 and M2 and a pore-forming domain H5. MbIRK3 shows around 65% amino acid identity with IRK1 and rbIRK2 and only 50% with ROMK1 and GIRK1. On the other hand, mbGIRK2 and mbGIRK3 are more similar to GIRK1 (60%) than to ROMK1 and IRK1 (50%). Northern blot analysis reveals that these three novel clones are mainly expressed in the brain. Xenopus oocytes injected with mbIRK3 and mbGIRK2 cRNAs display inward rectifier K(+)-selective currents very similar to IRK1 and GIRK1, respectively. As expected from the sequence homology, mbGIRK2 cRNA directs the expression of G-protein coupled inward rectifier K+ channels which has been observed through their functional coupling with co-expressed delta-opioid receptors. These results provide the first evidence that the GIRK family, as the IRK family, is composed of multiple genes with members specifically expressed in the nervous system.

Expression cloning in K+ transport defective yeast and distribution of HBP1, a new putative HMG transcriptional regulator

The rat HBP1 cDNA was cloned by its capacity to suppress the potassium transport-defective phenotype of mutant Saccharomyces cerevisiae cells. HBP1 cDNA encodes a 513 amino acids protein which, unexpectedly, does not share any homology with K+ transporters or K+ channels. However, a search in protein databases reveals that HBP1 contains a putative DNA-binding domain called HMG-box. Northern blot analysis shows that HBP1 is expressed in a variety of tissues and that in adipocyte and myogenic cell lines, its expression is directly related to differentiation. Taken together, the results suggest that the rat HBP1 is a new member of the HMG class of transcriptional regulators involved in cell differentiation pathways.

Multiple mRNA isoforms encoding the mouse cardiac Kv1-5 delayed rectifier K+ channel

The mouse Kv1-5 K+ channel cDNA has been cloned from heart. This channel was highly expressed in heart and, to a lesser extent, in other tissues, including brain and thymus. Two alternatively spliced isoforms were found. The longer form encoded a 602-amino acid protein, while in the short form (Kv1-5 delta 5'), the first 200 amino acids lying upstream the transmembrane segment S1 were deleted. RNase protection experiments showed that both Kv1-5 mRNA isoforms are present in the mouse tissues examined, the longer form being predominant. The short mRNA (Kv1-5 delta 5') arose by an unusual splicing event within the exonic sequence. An additional short cDNA clone (Kv1-5 delta 3') that codes for a carboxyl-terminal truncated protein has been isolated. The gene coding sequence contained a single exon and has been mapped on human chromosome 12 (p13) and on mouse chromosome 6 (band F). Expression in Xenopus oocytes revealed that the long (Kv1-5) and the amino-terminal deleted (Kv1-5 delta 5') isoforms elicited similar K+ currents with a drastically decreased efficacy for Kv1-5 delta 5'. The carboxyl-terminal truncated Kv1-5 delta 3' clone was not functional but inhibited the expression of the long isoform.

Expression of the transcripts initiated in the 62nd intron of the dystrophin gene

The pattern of expression of two distal transcripts initiated in the 62nd intron of the dystrophin gene was investigated under different circumstances; (i) during the development of different rat tissues these transcripts and Dp71, a protein encoded by one of them, increased with brain development and decreased with muscle development; (ii) in cultured glial and neuronal cells, the distal promoter was coactivated with tissue-specific upstream promoters, the muscle-type promoter in glial cells and the brain-type promoter in neuronal cells, which suggests that activity of the upstream promoter does not interfere with activity of the distal promoter; (iii) in lymphoblasts of DMD patients with various deletions of the dystrophin gene, the most distal of which included the 56th intron, the production of the distal transcript was not perturbed.

Expression of the dystrophin gene in cultured fibroblasts

The dystrophin whose defect is responsible for Duchenne and Becker muscular dystrophies is present in muscle, brain and cerebellum. We describe here the detection of dystrophin in human cultured skin fibroblasts, L809 cells and murine 3T6 cell line. Dystrophin transcripts initiated at the muscle specific first exon can also be amplified by cDNA-PCR from various fibroblastic cells. The expression of the dystrophin gene in fibroblasts could account for some abnormalities observed in patient's fibroblast cultures.

Distal transcript of the dystrophin gene initiated from an alternative first exon and encoding a 75-kDa protein widely distributed in nonmuscle tissues

A transcript generated by the distal part of the Duchenne Muscular Dystrophy (DMD) gene was initially detected in cells where the full size 14-kilobase (kb) messenger RNA is not found at a significant level. This transcript, approximately 4.5 kb long, corresponds to the cysteine-rich and carboxyl-terminal domains of dystrophin. It begins with a novel 80- to 100-nucleotide exon containing an ATG start site for a new coding sequence of 17 nucleotides in-frame with the consecutive dystrophin cDNA sequence from exon 63. This result suggests the existence of a third promoter that would be localized about 8 kilobases upstream from exon 63 of the DMD gene. The distal transcript is widely distributed but is absent in adult skeletal and myometrial muscle. It is much more abundant in fetal tissues. With an antibody directed against the dystrophin carboxyl terminus, the protein corresponding to this transcript was detected as a 70- to 75-kDa entity on Western blots. It was found in all tissues analyzed except in skeletal muscle. It was not found in lymphoblastoid cells from a Duchenne patient with a complete deletion of the dystrophin gene. The role and subcellular localization of this protein is not known. It may explain extramuscular symptoms exhibited by some Duchenne patients.

Positive and negative regulatory DNA elements including a CCArGG box are involved in the cell type-specific expression of the human muscle dystrophin gene

The muscle-specific promoter of the dystrophin gene is active in skeletal, cardiac, and smooth muscles and is specifically stimulated during differentiation of myoblasts into multinucleated myotubes. An 850-base pair (bp) DNA fragment upstream from the cap site is able to confer a partial muscle specificity to a reporter gene. The region between -850 and -140 bp includes nonspecific negative and positive regulatory sequences. A continuous stretch of 140 bp upstream from the cap site exhibits a striking conservation between rodents and human (93% homology) and still retains muscle preference of expression. It contains two putative binding sites for factors involved in regulation of other muscle-specific genes, a CCArGG box and an E box. This latter element, however, is unable to confer the ability to be transactivated by MyoD1 to the dystrophin promoter. The -140-bp promoter fragment exhibits antagonist effects contributed by one inhibiting sequence (nucleotide -140/-96), active in all cell types, and one activating region, from nucleotide -96 to the cap site, sufficient to confer a muscle preference of expression, in which the CCArGG box seems to play a major role.

Use of recombinant fusion proteins and monoclonal antibodies to define linear and discontinuous antigenic sites on the dengue virus envelope glycoprotein

Sixteen overlapping fragments of the dengue-2 virus envelope (E) protein, expressed as trpE-E fusion products in Escherichia coli, were used to map the epitopes defined by a panel of 20 monoclonal antibodies (MAbs) by immunoblotting. Using this technique, the amino acid sequence of six antigenic domains on the E protein was characterized. Nonneutralizing MAbs were found to define either linear-specific, subcomplex-specific (amino acids 22-58), and complex-specific (amino acids 304-332) epitopes or a subcomplex conformational-dependent epitope requiring the presence of two closely linked amino acid sequences from the E protein, 60-97 and 298-397. Neutralizing MAbs, however, defined either group-reactive epitopes present on two overlapping domains (amino acids 60-135; amino acids 60-205) or type-, subcomplex-, complex-, subgroup-, and group-specific determinants (amino acids 298-397). These neutralizing epitopes were all found to be dependent upon disulfide bridges. Our results suggest that the maintenance of a topographical arrangement of discontinuous antigenic domains in the flavivirus E-protein is necessary to induce neutralizing and protective antibodies.

Are cysteine-rich and COOH-terminal domains of dystrophin critical for sarcolemmal localization?

It has been hypothesized that the tight localization of dystrophin at the muscle membrane is carried out by its cysteine-rich and/or carboxyl domains. We report the results of biochemical and immunocytochemical investigations of dystrophin in muscle from a 1-yr-old patient with a large deletion that removes the distal part of the dystrophin gene, thus spanning the exons coding for the cysteine-rich and the carboxy-terminal domains, and extends beyond the glycerol kinase and congenital adrenal hypoplasia genes. Immunological analysis of muscle dystrophin shows that the deletion results in the production of a truncated, but stable, polypeptide correctly localized at the sarcolemma. These data indicate that neither the cysteine-rich domain, nor the carboxyl domain, are necessary for the appearance of normal dystrophin sarcolemmal localization.

Illegitimate transcription. Application to the analysis of truncated transcripts of the dystrophin gene in nonmuscle cultured cells from Duchenne and Becker patients

We have previously demonstrated that there is a low level of transcription of tissue-specific genes in every cell type. In this study, we have taken advantage of this phenomenon, called illegitimate transcription, to analyze the muscle-type dystrophin mRNA in easily accessible cells such as lymphoid cells, fibroblasts, and peripheral blood cells from Duchenne and Becker muscular dystrophies with known internal gene deletion. The results showed that, in the studied regions surrounding the deletions, processing of truncated transcripts is identical in specific (muscle tissue) and in nonspecific cells (lymphoid cells). In Becker cases with out-of-frame deletions, the already described alternatively spliced species found in muscle samples were also found in nonspecific cells. These results demonstrate that illegitimate transcripts are a bona fide version of tissue-specific mRNA, and that they represent a useful material to investigate the qualitative consequences of gene defects at the mRNA level.

Illegitimate (or ectopic) transcription proceeds through the usual promoters

Illegitimate transcription corresponds to the low level presence of specific transcripts in nonspecific cells. This phenomenon allows to analyse any tissue-specific disease transcript in any easily accessible cell. We demonstrate here that the start sites of transcription are the same in specific and non-specific cells, which indicates that illegitimate transcription is due to a low level activity of the normal promoter. In addition, it is possible to increase about 10 fold the abundance of illegitimate transcripts through the use of cycloheximide. This treatment should, therefore, facilitate detection and qualitative analysis of illegitimate transcripts.

A highly informative CACA repeat polymorphism upstream of the human dystrophin gene (DMD)

Images

Identification of dengue sequences by genomic amplification: rapid diagnosis of dengue virus serotypes in peripheral blood

Polymerase chain reaction (PCR) was developed for the in vitro amplification of dengue virus RNA via cDNA. A fraction of the N-terminus gene of the envelope protein in the four dengue serotypes was amplified using synthetic oligonucleotide primer pairs. Amplified products were cloned and used as dengue type-specific probes in gel electrophoresis and dot-blot hybridization. We detected and characterized dengue virus serotypes in blood samples by the three-step procedure DNA-PAH consisting in cDNA priming (P), DNA amplification (A) and hybridization (H) using specific non-radiolabelled probes. Our findings showed that DNA-PAH was more rapid and sensitive in the identification of the infecting serotype than the mosquito cell cultures. Moreover, the failure of cultures to detect virus particles in sera containing few copies of viral genome or anti-dengue antibodies justified the approach of DNA-PAH to the dengue identification in clinical specimens.