Search

Acquired demyelinating syndrome associated with myelin oligodendrocyte glycoprotein antibodies, named recently myelin oligodendrocyte glycoprotein-associated disease, represents >27% of this paediatric syndrome. Relapses occur in 40% of them, which may be associated with severe outcomes. Aiming to identify biomarker allowing to predict relapse, we measured both myelin oligodendrocyte glycoprotein antibodies and neurofilament light chain levels in blood samples of patients that are known to reflect axonal injuries in neurological diseases including demyelinating autoimmune disorders. Three groups of patients were selected: relapsing myelin oligodendrocyte glycoprotein-associated disease (n = 8), non-relapsing myelin oligodendrocyte glycoprotein-associated disease (n = 7) and control patients with non-inflammatory neurological diseases (n = 12). Neurofilament light chain concentrations were measured in plasma of these three groups of patients using the high-sensitivity single-molecule array method at onset of the disease and 6 months later. At onset of the disease, we found that levels of neurofilament light chain in blood of non-relapsing patients were significantly higher than in control patients (means: 98.36 ± 22.66 versus 12.47 ± 2.47 pg/mL, **P < 0.01, Kruskal–Wallis test). The mean neurofilament light chain value in relapsing patients (82.16 ± 38.41 pg/mL) was not significantly different from that in non-relapsing and in control patients. Plasma myelin oligodendrocyte glycoprotein antibody levels were 2.5-fold higher in relapsing than in non-relapsing patients without reaching significance (means: 15.26 ± 4.87 versus 5.96 ± 1.13; two-tailed Mann–Whitney U-test P = 0.119). Plasma neurofilament light chain correlated significantly with myelin oligodendrocyte glycoprotein antibody levels in relapsing (two-tailed Spearman r = 0.8, P = 0.0218) but not in non-relapsing (two-tailed Spearman r = 0.17, P = 0.71). Interestingly, the ratio of neurofilament light chain-to-myelin oligodendrocyte glycoprotein antibodies was significantly lower in relapsing than in non-relapsing patients (means: 5.19 ± 1.61 versus 21.87 ± 6.13; two-tailed Mann–Whitney U-test P = 0.014). These findings suggest that measuring both neurofilament light chain and myelin oligodendrocyte glycoprotein antibody levels in patients at onset of demyelinating disease could predict relapse of myelin oligodendrocyte glycoprotein-associated disease.

The authors describe quantitatively and qualitatively different forms of experimental autoimmune encephalitis (EAE) in cynomolgus macaques. They found that bacterial contaminants within recombinant human myelin oligodendrocyte glycoprotein seemed to aggravate disease evolution. They provide anatomopathological features of fulminant and progressive forms of EAE, allowing them to distinguish specific factors influencing the evolution of this model of autoimmune demyelinating disease.Experimental autoimmune encephalitis (EAE) is a well-recognized model for the study of human acquired demyelinating diseases (ADD), a group of inflammatory disorders of the central nervous system (CNS) characterized by inflammation, myelin loss, and neurological impairment of variable severity. In rodents, EAE is typically induced by active immunization with a combination of myelin-derived antigen and a strong adjuvant as complete Freund's adjuvant (CFA), containing components of the mycobacterial wall, while myelin antigen alone or associated with other bacterial components, as lipopolysaccharides (LPS), often fails to induce EAE. In contrast to this, EAE can be efficiently induced in non-human primates by immunization with the recombinant human myelin oligodendrocyte glycoprotein (rhMOG), produced in Escherichia coli (E. coli), purified and formulated with incomplete Freund's adjuvant (IFA), which lacks bacterial elements. Here, we provide evidence indicating how trace amounts of bacterial contaminants within rhMOG may influence the course and severity of EAE in the cynomolgus macaque immunized with rhMOG/IFA. The residual amount of E. coli contaminants, as detected with mass spectrometry within rhMOG protein stocks, were found to significantly modulate the severity of clinical, radiological, and histologic hallmarks of EAE in macaques. Indeed, animals receiving the purest rhMOG showed milder disease severity, increased numbers of remissions, and reduced brain damage. Histologically, these animals presented a wider diversity of lesion types, including changes in normal-appearing white matter and prephagocytic lesions. Non-human primates EAE model with milder histologic lesions reflect more accurately ADD and permits to study of the pathogenesis of disease initiation and progression.

Acquired demyelinating syndromes (ADS) are frequently associated with myelin oligodendrocytes glycoprotein (MOG) antibodies in children. Clinical phenotypes are heterogeneous and may delay the diagnosis, especially when they relapse and are atypical, mimicking diseases such as multiple sclerosis or neuromyelitis optica spectrum disorders . Here, we describe two children: one with a progressive cognitive and behavioral deterioration with seizures after only one relapse and the other with similar clinical impairments associated with multiple relapses. Brain magnetic resonance imaging revealed a subsequent progressive leukodystrophy-like lesion with diffuse bilateral white matter injuries in both patients. Cerebrospinal fluid analysis showed pleiocytosis, increased level of proteins with no oligoclonal bands. Metabolic and inflammatory blood markers were all negative. Brain biopsy was performed in the second child and nonspecific inflammatory lesions with no argument for histiocytosis or tumor were observed. Clinical and radiological stabilization were obtained after active immunotherapy. Retrospective analysis of anti-MOG antibodies in these two children was positive at the earlier stage of the disease and turned negative after treatment and during follow-up. Leukodystrophy-like ADS with anti-MOG-antibodies may display distinct progressive phenotype and have a severe neurological prognosis. Early diagnosis and appropriate treatment may improve outcome in these children.

AIM To describe cognitive abilities through the evaluation of academic difficulties in children with acute demyelinating syndromes (ADS) and myelin oligodendrocyte glycoprotein (MOG) antibodies. METHOD This was an observational, retrospective study of a French paediatric cohort that included children aged 18 years and younger. Clinical, biological, and imaging data were collected and academic outcome was measured. RESULTS Seventy-six children were included in the study with a mean (SD) follow-up of 4 years 7 months (6y 4mo). Median age at disease onset was 9 years 1 months (interquartile range=4y 7mo-13y 11mo; 36 females, 40 males). Thirty-six children relapsed and 20 had academic difficulties at the last follow-up. Academic difficulties, as well as deep grey matter and putaminal lesions (p=0.047 and p=0.006 respectively), were significantly more prevalent in children aged 10 years and younger (p=0.02). Using univariate binary regression analysis, we found that age at disease onset of 10 years and younger (odds ratio [OR] 3.72 [95% confidence interval {CI} 1.19-11.64]; p=0.024), acute disseminated encephalomyelitis at disease onset (OR 52.5 [95% CI 5.97-461.4]; p

Experimental autoimmune encephalitis (EAE) is a well-recognized model for the study of human acquired demyelinating diseases (ADD), a group of inflammatory disorders of the central nervous system (CNS) characterized by inflammation, myelin loss, and neurological impairment of variable severity. In rodents, EAE is typically induced by active immunization with a combination of myelin-derived antigen and a strong adjuvant as complete Freund's adjuvant (CFA), containing components of the mycobacterial wall, while myelin antigen alone or associated with other bacterial components, as lipopolysaccharides (LPS), often fails to induce EAE. In contrast to this, EAE can be efficiently induced in non-human primates by immunization with the recombinant human myelin oligodendrocyte glycoprotein (rhMOG), produced in Escherichia coli (E. coli), purified and formulated with incomplete Freund's adjuvant (IFA), which lacks bacterial elements. Here, we provide evidence indicating how trace amounts of bacterial contaminants within rhMOG may influence the course and severity of EAE in the cynomolgus macaque immunized with rhMOG/IFA. The residual amount of E. coli contaminants, as detected with mass spectrometry within rhMOG protein stocks, were found to significantly modulate the severity of clinical, radiological, and histologic hallmarks of EAE in macaques. Indeed, animals receiving the purest rhMOG showed milder disease severity, increased numbers of remissions, and reduced brain damage. Histologically, these animals presented a wider diversity of lesion types, including changes in normal-appearing white matter and prephagocytic lesions. Non-human primates EAE model with milder histologic lesions reflect more accurately ADD and permits to study of the pathogenesis of disease initiation and progression.

Background Autoantibodies against myelin oligodendrocyte glycoprotein (anti-MOG-Abs) occur in a majority of children with acquired demyelinating syndromes (ADS) and physiopathology is still under investigation. As cynomolgus macaques immunized with rhMOG, all develop an experimental autoimmune encephalomyelitis (EAE), we assessed relatedness between anti-MOG-Abs associated diseases in both species. Methods The study includes 27 children followed for ADS and nine macaques with rhMOG-induced EAE. MRI lesions, cytokines in blood, and CSF at onset of ADS or EAE, as well as histopathological features of brain lesions were compared. Results Twelve children with anti-MOG-Abs ADS (ADS MOG+) and nine macaques with EAE, presented increased IL-6 and G-CSF in the CSF, whereas no such signature was found in 15 ADS MOG−. Furthermore, IgG and C1q were associated to myelin and phagocytic cells in brains with EAE (n = 8) and in biopsies of ADS MOG+ (n = 2) but not ADS MOG− children (n = 1). Macaque brains also revealed prephagocytic lesions with IgG and C1q depositions but no leukocyte infiltration. Conclusions Children with ADS MOG+ and macaques with EAE induced with rhMOG, present a similar cytokine signature in the CSF and a comparable aspect of brain lesions indicating analogous pathophysiological processes. In EAE, prephagocytic lesions points at IgG as an initial effector of myelin attack. These results support the pertinence of modeling ADS MOG+ in non-human primates to apprehend the natural development of anti-MOG-associated disease, find markers of evolution, and above all explore the efficacy of targeted therapies to test primate-restricted molecules.

To study the effect of vaccination on tolerization to the myelin antigen MOG we used a macaque model of experimental autoimmune encephalitis (EAE) in which immunization with recombinant human myelin oligodendrocyte glycoprotein (rhMOG) elicits brain inflammation and demyelination mediated by MOG-specific autoreactive CD4+ T lymphocytes and anti-MOG IgG. For antigen targeting to tolerizing antigen presenting cells, we used a recombinant antibody directed to the Dendritic Cells (DC)-Asialoglycoprotein receptor (DC-ASGPR). The intradermal administration of an anti-DC-ASGPR-MOG fusion protein, but not a control anti-DC-ASGPR-PSA (prostate specific antigen) protein, protected monkeys committed to develop EAE. Although effective treatment did not modify anti-MOG IgG production, it prevented the CD4+ T lymphocyte activation and pro-inflammatory cytokine production. Moreover, animals treated with anti-DC-ASGPR-MOG experienced a rise of MOG-specific CD4+CD25+FOXP3+CD39+ regulatory T cells as well as a TGFβ1, TGFβ2 and IL-8 upsurge after rhMOG re-immunization. Our results indicate that the pathogenicity of autoantibodies directed to MOG is mitigated in the presence of MOG-specific regulatory lymphocytes. This vaccination scheme appears suitable to treat relapsing autoimmune diseases with identified autoantigens such as that harboring anti-MOG or anti-AQP4 autoantibodies.

Acquired demyelinating syndromes (ADS) in children evolve either as a monophasic disease diagnosed as acute demyelinating encephalomyelitis (ADEM), transverse myelitis (TM) or optic neuritis (ON), or a multiphasic one with several relapses most often leading to the diagnosis of multiple sclerosis (MS) or neuromyelitis optica (NMO). These neuroinflammatory disorders are increasingly associated with autoantibodies against proteins such as aquaporin-4 in rare instances, and more frequently against myelin oligodendrocyte glycoprotein (MOG). Recently, in adult NMO patients, C5a levels were shown to be elevated in cerebrospinal fluid (CSF) during acute exacerbation. We investigated the CSF levels of anaphylatoxins and pro-inflammatory cytokines, and plasma MOG antibodies in onset samples from children with ADS. Thirty four children presenting with a first episode of ADS, 17 with monophasic ADS (9 with ADEM, 4 with TM and 4 with ON) and 17 with MS, who had paired blood and CSF samples at onset were included and compared to 12 patients with other non-inflammatory neurological disorders (OND). Cytokines and anaphylatoxins in CSF were measured by Cytometric Bead Array immunoassay. MOG antibody titers in plasma were tested by flow cytometry using a stable cell line expressing full-length human MOG. We found a significant increase in C5a levels in the CSF of patients with monophasic ADS (n=17) compared to OND (n=12, p=0.0036) and to MS (n=17, p=0.0371). The C5a levels in MS were higher than in OND without reaching significance (p=0.2). CSF IL-6 levels were significantly increased in monophasic ADS compared to OND (p=0.0027) and to MS (p=0.0046). MOG antibody plasma levels were significantly higher in monophasic ADS (p

We sought to determine the pathway of HIV-1 entry into human astrocytes and the fate of HIV-1 by detecting viral DNA and GFP-tagged HIV-1 in HIV-1-infected primary astrocytes. Immunochemistry and FACS analysis were used to assess the expression of DC-SIGN in human purified cultures of astrocytes. HIV-1 LTR was detected by PCR in infected cultures of human embryonic astrocytes at their third passage. GFP-Vpr-labeled R5 tropic HIV-1 was used to infect astrocytes, and was followed by confocal microscopy. Forty percent of astrocytes expressed DC-SIGN at the membrane level. Viral DNA was detected 5 days after infection in human astrocytes, but not in the presence of anti-CCR5 and anti-DC-SIGN mAbs. T20, NH4Cl, and bafilomycin had no effect on viral DNA detection. We found that 67% of the fluorescent GFP-Vpr-labeled R5 tropic HIV-1 viruses were present in the endosomes of astrocytes at 24 h, but not in the presence of anti-CCR5 or DC-SIGN mAbs. Bafilomycin and NH(4)Cl each increased the amount of fluorescent HIV-1 detected outside endosomes. Titers of p24 remained low from day 1 to day 5 postinfection, in the presence or absence of NH4Cl. Astrocytes express DC-SIGN and HIV-1 penetrates into these cells through CCR5- and/or DCSIGN- mediated endocytosis, via a pH-dependent pathway, with a delayed reverse transcription after infection without productive infection.

Oxidative stress has been suggested to be an important mediator of dopaminergic cell death in Parkinson's disease (PD). We investigated the neuroprotective potential of Cu/Zn superoxide dismutase (SOD1) overexpression in the rat substantia nigra (SN) following adenovirus-mediated gene transfer. Human dopaminergic SK-N-SH cells were transduced with adenoviral vectors expressing either human SOD1 (Ad-SOD1) or β-galactosidase (Ad-βgal) before exposure to 1 mM of the 1-methyl-4-phenylpyridinium ion (MPP+). A strong neuroprotective effect of SOD1 gene transfer was observed in the SK-N-SH cells exposed to MPP+ compared with controls. Adult rats were then given unilateral injections of either Ad-SOD1 or Ad-βgal into the striatum, and MPP+ was administered 8 days later at the same location. Strong transgene expression was detected in the SN dopaminergic neurons, a consequence of retrograde axonal transport of the adenoviral particles. The amphetamine-induced rotational behavior of the rats was markedly lower in Ad-SOD1-injected rats than in control animals. Also, behavioral recovery significantly correlated with the number of tyrosine hydrolase-expressing neurons in the SN of the treated rats. These results are consistent with oxidative stress contributing to the MPP+-induced neurodegenerative process. They also indicate that SOD1 gene transfer into the nigrostriatal system may be a potential neuroprotective strategy for treating PD. © 2005 Wiley-Liss, Inc.

Amphotropic retroviruses with modified envelope displaying single-chain antibody fragment (scFv) directed against the c-Met receptor were recently generated and found to efficiently and selectively deliver genes into hepatocarcinoma cells. A large proportion of human gliomas also frequently overexpresses c-Met. We therefore explored the possibility of infecting glioma cells using such retroviruses bearing an scFv directed against c-Met. In one construct, a urokinase (uPA) cleavage site was inserted between the scFv and the envelope. We assessed the transduction by these chimeric viruses of a panel of seven human glioma cell lines that we characterized for their c-Met and uPA levels. We found that abundance of the c-Met receptor and viral infection were inversely correlated if we used the retrovirus displaying scFv directed against c-Met, suggesting that the chimeric virus binds preferentially to the c-Met receptor, resulting in virus sequestration. Addition of the uPA site between the scFv moiety and the envelope restored the infectivity of the virus, consistent with a "two-step" infection process: (1) virus binding to the c-Met receptor, (2) cleavage of the scFv moiety by uPA, enabling the virus to dissociate from c-Met and entry into the cells via the Pit-2 receptor. Our study has significant implications for the design of targeting strategies for gliomas expressing high levels of c-Met.

Background The caspase-3 gene is expressed at significantly lower levels in human hepatocellular carcinomas than in normal hepatocytes. Gene transfer technologies offer the possibility to restore caspase-3 gene expression. We explored the interest for cancer gene therapy of a constitutively active recombinant caspase-3 (RevCasp3) obtained by rearranging its subunits. Methods An amphotropic retroviral vector was used to express the RevCasp3 gene. HuH7 cells were infected 1 and 2 days after plating. Caspase-3 activity was measured every 24 h for the following 6 days. The level of poly (ADP-ribose) polymerase cleavage induced by caspase-3 was measured by Western blot. The percentage of apoptotic cells was estimated after Hoechst-acridine orange and TUNEL stainings. Results Caspase-3 activity significantly increased from days 4 to 7 after infection. Caspase-3 activity peaked on day 7, and was 5.4-fold higher in RevCasp3-transduced HuH7 cells than in control cells. Poly (ADP-ribose) polymerase cleavage was first detected 6 days after the first infection. Hoechst-acridine orange and TUNEL stainings showed that most infected HuH7 cells were apoptotic. Conclusions Apoptosis was selectively induced following infection of HuH7 cells with RevCasp3, demonstrating that retroviruses expressing RevCasp3 are of potential interest for the treatment of hepatocellular carcinomas and other tumour types. Copyright © 2004 John Wiley & Sons, Ltd.

To study the possibility that increasing striatal activity of aromatic L-amino acid decarboxylase (AADC; EC 4.1.1.28) can increase dopamine production in dopamine denervated striatum in response to L-3,4-dihydroxy-phenylalanine (L-DOPA) administration, we grafted Cos cells stably expressing the human AADC gene (Cos-haadc cells) into 6-hydroxydopamine denervated rat striatum. Before grafting, the catalytic activity of the enzyme was assessed in vitro via the generation of 14CO2 from L-[14C]DOPA. The Km value for L-DOPA in intact and disrupted cells was 0.60 and 0.56 mM, respectively. The cofactor, pyridoxal 5-phosphate, enhanced enzymatic activity with maximal effect at 0.1 mM. The pH optimum for enzyme activity was 6.8. Grafting Cos-haadc cells into denervated rat striatum enhanced striatal dopamine levels measured after systemic administration of L-DOPA. When measured 2 h after L-DOPA administration, the mean dopamine level in the striata of Cos-haadc-grafted animals was 2 micrograms/g of tissue, representing 31% of normal striatal dopamine concentration. The mean dopamine concentration in the striata grafted with untransfected Cos cells (Cos-ut cells) was 1 microgram/g. At 6-8 h after L-DOPA administration, striatal dopamine content in the Cos-haadc-grafted animals was 0.67 microgram/g of tissue weight, representing 9% of intact striatum dopamine content. By contrast, the average dopamine content in the Cos-ut-grafted animals was undetectable. These findings demonstrate that enhancing striatal AADC activity can improve dopamine bioformation in response to systemically administered L-DOPA.

Adenovirus is an efficient vector for neuronal gene therapy due to its ability to infect post-mitotic cells, its high efficacy of cell transduction and its low pathogenicity. Recombinant adenoviruses encoding for therapeutical agents can be delivered in vivo after direct intracerebral injection into specific brain areas. They can be transported in a retrograde manner from the injection site to the projection cell bodies offering promising applications for the specific targeting of selected neuronal populations not easily accessible by direct injection, such as the motor neurons in the spinal cord. Adenoviral vectors are also efficient tools for the ex vivo gene therapy, that is, the genetical modification of cells prior to their transplantation into the nervous system. Recently, the efficacy of the adenovirus as a gene vector system has been demonstrated in several models of neurodegenerative diseases including Parkinson's disease (PD) and motor neuron diseases. In rat models of PD, adenoviruses encoding for either tyrosine hydroxylase, superoxide dismutase or glial-derived neurotrophic factor improved the survival and the functional efficacy of dopaminergic cells. Similarly, the intramuscular injection of an adenovirus encoding for neurotrophin-3 had substantial therapeutic effects in a mutant mouse model of motor neuron degenerative disease. However, although adenoviruses are highly attractive for neuronal gene transfer, they can trigger a strong inflammatory reaction leading in particular to the destruction of infected cells. The recent development of new generations of adenoviral vectors could shed light on the nature of the immune reaction caused by adenoviral vectors in the brain. The use of these new vectors, combined with that of neurospecific and regulatable promoters, should improve adenovirus gene transfer into the central nervous system.

Glial-cell-line-derived neurotrophic factor (GDNF) is a potent neurotrophic factor for adult nigral dopamine neurons in vivo . GDNF has both protective and restorative effects on the nigro-striatal dopaminergic (DA) system in animal models of Parkinson disease. Appropriate administration of this factor is essential for the success of its clinical application. Since it cannot cross the blood–brain barrier, a gene transfer method may be appropriate for delivery of the trophic factor to DA cells. We have constructed a recombinant adenovirus (Ad) encoding GDNF and injected it into rat striatum to make use of its ability to infect neurons and to be retrogradely transported by DA neurons. Ad-GDNF was found to drive production of large amounts of GDNF, as quantified by ELISA. The GDNF produced after gene transfer was biologically active: it increased the survival and differentiation of DA neurons in vitro . To test the efficacy of the Ad-mediated GDNF gene transfer in vivo , we used a progressive lesion model of Parkinson disease. Rats received injections unilaterally into their striatum first of Ad and then 6 days later of 6-hydroxydopamine. We found that mesencephalic nigral dopamine neurons of animals treated with the Ad-GDNF were protected, whereas those of animals treated with the Ad-β-galactosidase were not. This protection was associated with a difference in motor function: amphetamine-induced turning was much lower in animals that received the Ad-GDNF than in the animals that received Ad-β-galactosidase. This finding may have implications for the development of a treatment for Parkinson disease based on the use of neurotrophic factors.

Gene therapy is a potentially powerful approach to the treatment of neurological diseases. The discovery of neurotrophic factors inhibiting neurodegenerative processes and the isolation of genes encoding neurotransmitter synthesizing enzymes provide the basis for current gene therapy strategies for Parkinson's disease. Adenovirus vectors have been shown recently to allow efficient gene transfer to the brain. One of the advantages of recombinant adenovirus is that it can transduce both quiescent and actively dividing cells. Thus expression of transgenes in neurons using adenoviruses is possible after either direct in vivo gene transfer or ex vivo gene transfer. In vivo gene transfer, consisting of the direct intracerebral injection of genetic material, is a novel method that is particularly efficient with the adenoviral vector. Ex vivo gene transfer, combining gene transduction with intracerebral transplantation, is a way to improve the classical grafts which are limited by poor cell survival in Parkinson's disease. Probably because the brain is a partially immunologically privileged site, the expression of adenoviral vectors persists for several months with little inflammation. Recombinant adenoviruses are currently being improved, particularly by inactivating viral genes controlling the expression of immunodominant viral proteins.

Transplantations for neurological disorders are limited by the supply of human fetal tissue. To generate larger numbers of cells of appropriate phenotype, we investigated whether human neural progenitors expanded in vitro could be modified with recombinant adenoviruses. Strong expression of beta-galactosidase was obtained in vitro. Two or three weeks after transplantation of engineered cells to the rat brain, we observed a small percentage of surviving neuroblasts strongly expressing beta-galactosidase in four out of 13 rats. Thus human precursor cells that have been genetically modified using adenoviruses are a promising tool for ex vivo gene therapy of neurodegenerative diseases.

Direct intracerebral gene transfer to neural cells has been demonstrated with recombinant adenovirus encoding beta-galactosidase. To explore the potential of recombinant adenovirus for the therapy of neurological disease we constructed a recombinant adenovirus encoding tyrosine hydroxylase and optimized intracerebral injection to express the gene in the striatum of unilaterally denervated rats. These animals have dopamine depletion in their lesioned striatum, causing a rotation asymmetry induced by apomorphine. One and two weeks after intracerebral injection this sensorimotor asymmetry was decreased by the adenovirus encoding tyrosine hydroxylase and not by a control adenovirus encoding beta-galactosidase. Histological analysis showed that tyrosine hydroxylase was preferentially expressed in astrocytes.

A mouse line transgenic for nerve growth factor (NGF) was developed using the mouse prepro-NGF cDNA inserted within a plasmid containing the proximal region (−10 to −550 bp) of the c- fos promoter and the transcription termination and polyadenylation signals of the rabbit β-globin gene. No significant modification of gross behavior or central nervous system anatomy was detected in adult animals as assessed by immunohistochemistry and in situ hybridization for NGF and choline acetyltransferase. The expression of the transgene and the possible regulation of its expression by agents acting on the promoter were investigated in vitro. Despite the presence of an additional pool of NGF mRNA specific to the transgene, basal levels of NGF in the supernatant of transgenic astrocytes were similar to normal ones. On the other hand, transgenic neurons spontaneously synthesized and released levels of NGF two to three times higher than normal neurons, while mRNA levels were barely detectable by conventional Northern blotting. The tissue-specificity of NGF expression was respected, with higher levels in hippocampal than neocortical neurons. Increases of NGF mRNA by agents acting on the promoter could be observed in normal and transgenic astrocytes only after inhibition of the protein synthesis by cycloheximide, suggesting a similar rapid turnover of normal and transgenic transcripts. Cyclic AMP agonists specifically increased the secretion of NGF protein by transgenic astrocytes and neurons, while activators of the protein kinase C had a similar effect on transgenic and normal cells. Differences between amounts of NGF secreted by neurons and astrocytes with regards to their respective content in mRNA suggest that transgenic transcripts are subject to normal cell- and tissue-specific post-transcriptional regulations. Agents acting on the c- fos promoter through the protein kinase C or cyclic AMP routes differentially increased the secretion of NGF by transgenic astrocytes or neurons, supporting this hypothesis.

The therapeutic potential of gene transfer into somatic cells is increasingly considered in human diseases. Strategies are conceived to treat some diseases in which the delivery of a gene product can counteract a pathological process. There are three types of disease that form the target of these new potential treatments. In the first place, inherited gene defects; secondly, cell deficiencies leading to a lack of a gene product controlling a life-critical physiological function and thirdly, acquired diseases of tumour and/or viral origin. Such therapeutic tools must be evaluated in terms of efficacy and safety. They also have to be improved with respect to their effects in the perspective of a viable therapy for patients. This goal requires a strategy based on progressive steps toward efficient and safe clinical applications. This review will first survey the history of gene transfer technology and then describe the tools which are mainly being studied at present and how they are being developed to obtain a significant improvement of function.

We have constructed a recombinant retrovirus that expresses choline acetyltransferase (ChAT) by placing the porcine enzyme cDNA under the control of the 5′ long terminal repeat of the retroviral vector pMMuLV. Using retrovirus-mediated gene transfer, we have expressed ChAT in astroglial (STR-SVLT) and neuroendocrine (RIN) cell lines. Both genetically modified cell types synthesize acetylcholine (ACh). ACh is also present in the culture medium at a low concentration relative to that found in the modified cells. This result suggests that the synthesized ACh is retained within the cells and released by these two cell types. Release of ACh is not increased in the presence of the calcium ionophore A23187 or by depolarizing concentrations of potassium in either STR-SVLT or in RIN cells. The implications of these studies for understanding ACh release mechanisms are discussed.

International audience; Human immunodeficiency virus-1 (HIV-1) invades the brain early in infection and may cause HIV-associated dementia (HAD), which is characterized by reactive astrocytes, and macrophage and T-cell infiltrates. HIV-1 Tat protein is thought to contribute to HAD by transactivating host genes, such as that encoding monocyte chemoattractant protein-1 (MCP-1/CCL2), although its mechanisms of action are not fully understood. We investigated the molecular pathways involved in Tat-induced MCP-1/CCL2 gene expression in human astrocytes. We found that Tat induced MCP-1/CCL2 synthesis in human astrocytes infected with a lentivirus carrying the gene encoding Tat or treated with a biologically active synthetic Tat protein. The induction of MCP-1/CCL2 was independent of the nuclear factor kappaB (NF-kappaB) classical pathway, but was significantly inhibited by specific cyclin-dependent kinase 9 (cdk9) inhibitors, such as a dominant-negative mutant or siRNA. By contrast, broader-spectrum cdk inhibitors, such as roscovitine, 5,6-dichloro-1-beta-d-ribofuranosylbenzimidazole (DRB), and flavopiridol, inhibited MCP-1/CCL2 induction by Tat. We also analyzed the effects of roscovitine, DRB, and flavopiridol on Tat-induced HIV-1 long terminal repeat (LTR) expression following the infection of astrocytes and HeLa cells. Astrocytes showed no inhibition by roscovitine, 59% inhibition by DRB, and 80% inhibition by flavopiridol. In control HeLa cells, high levels of inhibition were observed with roscovitine, DRB, and flavopiridol. We have ascertained the direct implication of cdk9 in Tat-induced MCP-1 expression by performing ChIP assay. These results demonstrate that cdk9 is involved in Tat-induced HIV-1 LTR, MCP-1/CCL2 gene expression.

We report the cloning of KIN17 cDNA, 1414 bp long with an ORF of 391 residues showing a zinc finger and nuclear localization signals. By recloning the cDNA into an appropriate vector, we produced kin17 protein in E. coli, purified it partially and shown that kin17 protein binds to double-stranded DNA. The KIN17 gene was localized by cytogenetic mapping in mouse chromosome 2, band A. Genomic sequences homologous to KIN17 cDNA were detected also in rat and human DNAs. KIN17 mRNA is highly expressed in rodent transformed AtT-20 neuroendocrine cells whereas it can be detected only in the total RNA of mouse embryos and various normal adult tissues by reverse transcription and PCR amplification. The mouse nuclear kin17 protein was identified by a local small structural similarity with E.coli recA protein. Kin17 and recA have only 39 amino acid residues in a region that might be involved in DNA-binding.

Fibroblastic 3T3 and endocrine RIN cells were genetically modified by infection with a recombinant retrovirus encoding the form I of human tyrosine hydroxylase (TH) and selection in tyrosine-free medium. These cells were grafted to rats unilaterally lesioned with 6-hydroxy-dopamine. Both cell types survived implantation into the striatum, expressed TH immunoreactivity, and as assessed by microdialysis 8-9 days after implantation, secreted high amounts of DOPA and/or dopamine into the surrounding host striatum. The modified 3T3 cells secreted large amounts of DOPA that was efficiently decarboxylated to dopamine by the host striatal tissue; the newly synthesized dopamine was stored only to a limited extent in the denervated striatum. The modified RIN cells synthesized dopamine that was stored intracellularly and released in a regulated fashion. The grafted DOPA-secreting cells produced 4-5 times higher extracellular dopamine levels than the dopamine-secreting cells, and they were more efficient in reducing apomorphine-induced rotation. No effect was observed with either cell type on amphetamine-induced turning behavior.

Cell lines in which tyrosine hydroxylase was introduced either by infection or transfection were used in grafting experiments in a rat model of Parkinson's disease obtained by unilateral lesion of the substantia nigra. A neuroblastoma NS20 Y cell line which synthesizes only l-dopa and a neuroendocrine AtT-20 cell line which produces dopamine were obtained. They were grafted into denervated striata and their ability to compensate for the dopaminergic deficit was studied. Both modified cell types displayed a rapid partial reversal of apomorphine-induced turning behaviour. No effect was observed with the control unmodified cell lines. We discuss the usefulness of engineered cell lines to address the fundamental issues in grafting experiments and more particularly in the therapy of Parkinson's disease.

Tumors frequently express urokinase (uPA) receptor (uPAR). To investigate whether uPAR can efficiently target cancerous cells using amphotropic retroviral vectors, we generated a retrovirus displaying the amino-terminal fragment (ATF) of uPA as an N-terminal extension of viral envelope protein. We also made use of a “two-step strategy” by inserting a uPA cleavage site between the ATF moiety and the envelope. We measured the ability of ATF-bearing chimeric envelopes to infect huPAR-overexpressing Madin–Darby canine kidney (MDCK) and control MDCK II cells. The ATF-viruses infected both MDCK cell lines with an equivalent efficiency, suggesting that the chimeric viruses were not sequestered by uPAR and infect cells preferentially via the Pit-2 receptor. The addition of a uPA cleavage site increased the infection level of huPAR-MDCK cells by 2-fold when uPA was present in the infection medium. Surprisingly, ATF-env viruses infected huPAR-MDCK cells 5.5-fold more efficiently in the presence of exogenous uPA. This stimulatory effect of uPA on infection of huPAR-MDCK cells by the ATF-env virus was completely abolished by methyl-β-cyclodextrin, suggesting that this effect involves the caveolar endocytosis pathway.

A proliferation-inducing ligand (APRIL) of the tumour necrosis factor (TNF) family is produced in small amounts in many tissues and more abundantly in tumours. APRIL has been reported to promote cell growth in vivo and in vitro. It was recently shown that the production of APRIL in some glioblastoma cell lines does not lead to an increase in cell growth. In this study, we investigated the production of APRIL and its ability to increase the proliferation of eight human glioblastoma cell lines. We found that APRIL was produced in the eight human glioblastoma cell lines tested but not in the normal embryonic astrocyte counterparts of glioblastomas. Flow cytometry demonstrated the presence of a specific APRIL-binding receptor on the cell surface in all the glioblastoma cell lines tested. This receptor was also present on normal embryonic and adult astrocytes and embryonic neural progenitor cells. Moreover, the addition of recombinant human APRIL resulted in an increase in proliferation rate of normal adult astrocytes and in four of eight cell lines tested. Addition of the soluble recombinant TNF-receptor-homologue B-cell maturation (BCMA) chimeric protein, which binds APRIL, confirmed the involvement of APRIL in the growth of malignant glioblastoma cell lines.

Background —To define the role of metalloproteinases (MMPs) in the development of lipid-rich atherosclerotic lesions in relation to the balance between proteolytic and antiproteolytic activities, we investigated the impact of adenovirus-mediated elevation in the circulating levels of human tissue inhibitor of MMP (TIMP-1) in atherosclerosis-susceptible apolipoprotein E–deficient (apoE −/− ) mice. Methods and Results —Infusion of apoE −/− mice fed a lipid-rich diet with rAd.RSV.TIMP-1 (1×10 11 viral particles) resulted in high hepatic expression of TIMP-1. At 2 weeks after injection, plasma TIMP-1 levels ranged from 7 to 24 μg/mL (mean 14.8±6.8). Marked overexpression of TIMP-1 was transient, with levels of TIMP-1 decreasing to 2.5 to 8 μg/mL (mean 4.3±2.1) at 4 weeks. Plasma lipid and lipoprotein levels in mice treated with rAd.RSV.TIMP-1 were similar to those treated with rAd.RSV.βGal. However, rAd.RSV.TIMP-1–infused mice displayed a marked reduction (≈32%; P 2 ×10 3 ; n=12 sections from 4 animals) as compared with rAd.RSV.βGal-infused mice (750±182 μm 2 ×10 3 ; n=12 sections from 4 animals). Similarly, marked reduction in macrophage deposition as well as MMP-2, MMP-3, and MMP-13 antigens was observed. Conclusions —Histological and immunohistologic analyses of atherosclerotic lesions revealed increases in collagen, elastin, and smooth muscle α-actin content in mice treated with rAd.RSV.TIMP-1. These qualitative and quantitative features were the consequence of TIMP-1 infiltration from plasma to arterial intima, as immunohistochemical analyses revealed an abundance of TIMP-1 specifically in lesions of rAd.RSV.TIMP-1–treated mice.

We have developed a novel Schwann cell line, SCTM41, derived from postnatal sciatic nerve cultures and have stably transfected a clone with a rat glial cell line-derived neurotrophic factor (GDNF) construct. Coculture with this GDNF-secreting clone enhancesin vitrosurvival and fiber growth of embryonic dopaminergic neurons. In the rat unilateral 6-OHDA lesion model of Parkinson’s disease, we have therefore made cografts of these cells with embryonic day 14 ventral mesencephalic grafts and assayed for effects on dopaminergic cell survival and process outgrowth. We show that cografts of GDNF-secreting Schwann cell lines improve the survival of intrastriatal embryonic dopaminergic neuronal grafts and improve neurite outgrowth into the host neuropil but have no additional effect on amphetamine-induced rotation. We next looked to see whether bridge grafts of GDNF-secreting SCTM41 cells would promote the growth of axons to their striatal targets from dopaminergic neurons implanted orthotopically into the 6-OHDA-lesioned substantia nigra. We show that such bridge grafts increase the survival of implanted embryonic dopaminergic neurons and promote the growth of axons through the grafts to the striatum.

Gene therapy is a potentially powerful approach to the treatment of neurological diseases. The discovery of neurotrophic factors inhibiting neurodegenerative processes and neurotransmitter-synthesizing enzymes provides the basis for current gene therapy strategies for Parkinson's disease. Genes can be transferred by viral or nonviral vectors. Of the various possible vectors, recombinant retroviruses are the most efficient for genetic modification of cells in vitro that can thereafter be used for transplantation (ex vivo gene therapy approach). Recently, in vivo gene transfer to the brain has been developed using adenovirus vectors. One of the advantages of recombinant adenovirus is that it can transduced both quiescent and actively dividing cells, thereby allowing both direct in vivo gene transfer and ex vivo gene transfer to neural cells. Probably because the brain is partially protected from the immune system, the expression of adenoviral vectors persists for several months with little inflammation. Novel therapeutic tools, such as vectors for gene therapy have to be evaluated in terms of efficacy and safety for future clinical trials. These vectors still need to be improved to allow long-term and possibly regulatable expression of the transgene.

A cerebral endothelial immortalized cell line was used in transplantation experiments to deliver gene products to the adult rat brain. Survival of grafted cells was observed for at least 1 year, without any sign of tumor formation. When genetically modified to express bacterial beta-galactosidase and transplanted into the striatum, these cells were shown, by light and electron microscope analysis, to integrate into the host brain parenchyma and microvasculature. Following implantation into the striatum and nucleus basalis of adult rats, endothelial cells engineered to secrete mouse beta-nerve growth factor (NGF) induced the formation of a dense network of low-affinity NGF receptor-expressing fibers near the implantation sites. This biological response was observed from 3 to 8 weeks after engraftment. The present study establishes the cerebral endothelial cell as an efficient vector for gene transfer to the central nervous system.

Astrocytes secreting high levels of L-3,4-dihydroxyphenylalanine (DOPA) have been generated by retrovirus-mediated transfer of the human tyrosine hydroxylase (TH) gene. Immature astrocytes obtained from prenatal rat brain were cocultured with TH virus producing psi-2 cells that had been pretreated with the mitosis inhibitor mitomycin-C. During the first week of coculture DOPA production gradually increased to reach a plateau after 7-9 days. At this time point virtually all cells were GFAP positive and over 80% of them expressed TH. DOPA production in the transduced astrocytes was largely independent of exogenous cofactor, and DOPA release into the medium was not influenced by addition of either KCl or tetrodotoxin or by removal of Ca2+ from the culture medium, indicating that the newly synthesized DOPA was constitutively released from the cells. Transplantation of the TH-transduced astrocytes to the striatum in unilaterally 6-hydroxydopamine lesioned rats reduced apomorphine-induced turning by about 50% at 2 weeks postgrafting. Microscopic analysis revealed that the transduced astrocytes survived very well after transplantation and that some of the grafted cells had migrated out, partly along blood vessels, into the surrounding striatum. TH expression was observed in cells with both the appearance of mature GFAP-positive astrocytes, as well as in more immature-looking cells. However, only a few percent of all transplanted cells maintained significant expression of the transgene, as determined by TH immuno-histochemistry. The results show that primary astrocytes may be highly useful as gene carriers for ex vivo gene therapy in the CNS. With future improvement in the gene transduction procedure for more efficient, sustained expression of the TH transgene in vivo, genetically engineered DOPA-producing astrocytes hold great promise as a tool to explore the potential of ex vivo gene therapy in Parkinson's disease.

Publisher Summary This chapter describes the application of adenovirus to transfer genes into the central nervous system for treatment of neurodegenerative disorders. The recombinant adenoviruses commonly used are derived from type 5 adenovirus, which has never been found to be associated with tumoral pathology in humans. The probability of integration of the adenoviral DNA into the genome of the host cell is low. The virus life cycle includes attachment to cell-specific receptors, internalization into endosomes, release in the cytoplasm by endosmolysis, and translocation to the nucleus, where activation of the early regions occurs followed by DNA replication and activation of the late regions. The combined use of human neural progenitors amplified in vitro and adenoviruses may potentially allow the treatment of neurodegenerative disorders. Enough cells expressing a gene of interest, encoding atrophic factor or a neurotransmitter synthesizing enzyme, could be obtained from a single fetus to graft several patients. It is found that the in vitro step to amplify the cells allows testing for the absence in the fetal tissue of contaminating agents, such as viruses, and thereby results in improved safety. The immunological aspects and improvement of the recombinant adenoviral tool are also elaborated.

In situ hybridization experiments, using oligodeoxyribonucleotides specific for the two major expressed human tyrosine hydroxylase mRNAs, were performed on human brain sections at the level of the mesencephalon. The specificity of the probes was ascertained by Northern blot experiments carried out with independently in vitro synthesized human tyrosine hydroxylase mRNAs. For in situ hybridization experiments, oligodeoxyribonucleotides were labelled with nucleotides tagged with digoxigenin or biotin molecules. The hybridized oligonucleotides were detected by antibodies coupled with peroxidase and alkaline phosphatase enzymes, which yield, with appropriate substrates, brown and purple products, respectively. The simultaneous detection of the two mRNAs with digoxigeninated and biotinylated probes revealed that these two mRNAs are co-expressed in single cells. The purple product obtained with alkaline phosphatase exhibits a discrete distribution within the dopaminergic cells suggesting these mRNAs are associated with sub-cellular structures. Finally, a heterogeneity in the intensity of the labelling of reactive cells with both probes was visualized as well as the expression of the two mRNA species in neurites.

The single human tyrosine hydroxylase (TH) gene generates four different mRNA species through alternative splicing events. TH-1 and TH-2 mRNAs are expressed mostly in the brain. We have produced large amounts of the corresponding proteins in Escherichia coli to analyze their respective molecular characteristics. The polypeptides have molecular weights similar to those of TH expressed in Xenopus oocytes and react with antibodies to TH. The two isoforms were purified with a purity of 90% using a three-step procedure. The phosphorylation sites have been determined in the two isoforms after labeling with [gamma-32P]ATP in the presence of cAMP-dependent protein kinase (PKA) or calmodulin-dependent protein kinase II (CaM-PK II). In both isoforms, Ser-40 was found to be phosphorylated by PKA, and Ser-19 and Ser-40 were found to be phosphorylated by CaM-PK II. The putative phosphorylation site generated by alternative splicing (Ser-31) was phosphorylated specifically by CaM-PK II in TH-2 only. The kinetic properties of the two isoforms in the presence of various concentrations of the substrate (tyrosine) and of the natural cofactor [6R)-tetrahydrobiopterin) were also analyzed. TH produced in E. coli is unphosphorylated but nevertheless active. At 50 microM tyrosine and 300 microM (6R)-tetrahydrobiopterin, the specific activities of TH-1 and TH-2 are 1300 and 620 nmol of dihydroxyphenylalanine/min/mg, respectively. Phosphorylation of TH-1 and TH-2 by PKA activates both isoenzymes as shown by the increase in the affinity for the cofactor. No changes in kinetic parameters of the isoenzymes were observed after phosphorylation by CaM-PK II. Dopamine was found to inhibit both TH isoenzymes to the same extent as shown by their similar Ki values for dopamine. These values were increased after phosphorylation of each enzyme by PKA. Unlike TH-1, phosphorylation of TH-2 by CaM-PK II resulted in an increase of the Ki value for dopamine. This property may be related to the presence of the additional phosphorylated residue in TH-2 isoform.

Publisher Summary This chapter describes the exogeneous expression of tyrosine hydroxylase (TH) in the various cell lines of neuronal and nonneuronal origins. In an experiment described in the chapter, rat and human TH complementary DNAs (cDNAs) have been introduced in a neuroblastoma (NS20 Y), a fibroblast (NIH 3T3), and neuroendocrine (AtT-20) cell lines. The exogeneous expression of TH yields only L-dopa in the two former cell lines, while the latter releases dopamine (DA) in a regulated fashion. Grafting of both NS20 Y, which synthesizes L-dopa, and AtT-20, which produces DA, promote a partial reversal of the lesion-induced motor asymmetry within a few days. The implications of this approach to address fundamental issues in the mechanism of neurotransmitter release and grafting experiments are discussed in the chapter. The use of this approach allows the further analysis of mechanisms underlying functional recovery in the animal models of Parkinson's disease. Important issues such as (1) the role of the synthesis of neurotransmitters and peptides, (2) the importance of synapse formation, and (3) the involvement of growth factors and cell adhesion molecules may be addressed.

Embryonic striatal neurons transplanted into marmoset brains become integrated into host tissue and induce functional recovery in a primate model of Huntington's disease (pages 727−729).

Intracerebral transplantation of embryonic dopaminergic cells have shown, in particular, that local striatal restoration of dopaminergic neurotransmission may compensate for sensori-motor impairment associated with the nigro-striatal degeneration in animal models of Parkinson' disease. We are studying gene transfer as a tool to express the neurotransmitter synthesizing enzyme tyrosine hydroxylase (TH) locally in the striatum. We are exploring both the retroviral system and the adenoviral one, with the aim of comparing their efficiency and safety. Initially, we have constructed two human TH recombinant retroviruses using either the Long Terminal Repeat (LTR) from the Moloney murine leukemia virus or the immediate early promoter from the cytomegalovirus to drive the expression of TH into primary cultures of embryonic striatal cells. Retroviruses only permit the genetic modification through incorporation into chromosomal DNA of dividing cells. We therefore made use of the ability of glial cells to divide in the presence of foetal calf serum or growth factors to allow their infection in vitro by the recombinant retroviruses. We could thereby modify up to 100% of glial cells as characterized by their endogenous GFAP, an astroglial marker, as well as the exogenous TH protein expression. The genetically modified astroglial cells were found to release DOPA in a constitutive manner in vitro. The functional capacity of the TH-expressing astroglial cells as well as their survival properties were studied after grafting to striatum of rats with 6-hydroxydopamine lesions. Two weeks after transplantation, apomorphine-induced turning was reduced by mean of 60%. TH immunoreactive cells were observed in the denervated striatum three weeks postgrafting. The adenovirus allows the infection of dividing as well as of non-dividing cells such as neurons, resulting in the introduction of recombinant genes in an episomal form. The relative efficiency of this system with that of the retrovirus system will be discussed.

Little is known about the molecular events mediating neurotransmitter release, a crucial step in synaptic transmission. In this paper, the biosynthesis and release of L-beta-3,4-dihydroxyphenylalanine (L-DOPA) and dopamine were analyzed in three heterologous cell lines after retroviral-mediated gene transfer of tyrosine hydroxylase (EC 1.14.16.2), the rate-limiting enzyme in catecholamine synthesis. A recombinant retrovirus encoding human tyrosine hydroxylase type I as well as neomycin-resistance gene was used to infect a fibroblast (NIH 3T3), a neuroblastoma (NS20 Y), and a neuroendocrine (AtT-20) cell line. After selection in the presence of neomycin and in tyrosine-free medium, high levels of exogenous tyrosine hydroxylase activity were detected in extracts of the three cell lines. High-performance liquid chromatography of cell extracts and culture supernatants confirmed that the three cell lines hydroxylated tyrosine to form L-DOPA and released this metabolite into the culture medium. Interestingly, the neuroendocrine cell line AtT-20 synthesized not only L-DOPA but also dopamine. Evoked secretion studies established that AtT-20 cells released the transmitter upon depolarization in a regulated, calcium-dependent way. We discuss the implication of this approach for the analyses of neurotransmitter release as well as in the context of degenerative disorders such as Parkinson disease.

Tyrosine hydroxylase, the rate limiting enzyme in the biosynthesis of catecholamine, is a tetramer composed of four subunits of the same molecular mass. A full length cDNA clone encoding tyrosine hydroxylase has been inserted into the SP6 expression system. Translation of the corresponding RNA in Xenopus oocyte results in enzymatic activity, demonstrating that a single gene contains all the necessary genetic information to code for a functional enzyme. The potential of this system in the analysis of posttranslational tyrosine hydroxylase modifications is discussed.

A single human gene has been described to encode multiple tyrosine hydroxylase (TH) mRNAs. The study of this variation has been extended by S1 mapping experiments and by analysis of the 5' region of the TH gene. Four different mRNAs were found to originate solely from alternative splicing of two exons. Comparison of the 5' flanking regions of human and rat genes discloses several highly conserved segments, likely to play an important role in the regulation of TH gene expression.

A single human tyrosine hydroxylase (HTH) gene has been shown previously to generate four species of mRNA by alternative splicing. The four different HTH mRNAs were independently synthesized in vitro, using the SP6 transcription system. Each of these mRNA species was able to direct the synthesis of an active form of TH following injection into Xenopus oocytes. Quantitation of synthesized HTH polypeptides allowed the determination of the relative specific activity of each individual HTH form. A significant difference in specific activity was found between each form, suggesting that alternative splicing may play a role in regulating HTH activity in vivo.

Glioma and renal cell carcinoma (RCC) cells express high affinity interleukin 13 (IL13) binding sites, but only RCC cell proliferation was inhibited by IL13. Both of these two cell types are IL2-receptor (gamma)c chain-negative. We thus used these cell models to investigate the patterns of expression of IL13Ralpha1, IL13Ralpha2, and IL4Ralpha chains and the role of IL13Ralpha2 in the response to IL13. Using new specific antibodies and flow cytometry, we observed a similar surface expression of IL4Ralpha and IL13Ralpha1 chains in most RCC and glioma cells, whereas IL13Ralpha2 was only present on five of six glioma cell lines. In all glioma cell lines, the amount of IL13Ralpha2 expression was 10 to 30 times higher than that of the two other chains. Although there was no surface or intracellular expression of IL13Ralpha2, its mRNA was detected in three of seven RCC cell lines. The expression on RCC cells of IL13Ralpha2 mRNA and/or that of high-affinity IL13 binding sites is not sufficient to predict IL13Ralpha2 protein expression. Blocking experiments showed that IL4 and IL13 strongly inhibited RCC cell proliferation through a unique receptor composed of IL4Ralpha and IL13Ralpha1 chains. Using RCC cells stably transfected with IL13Ralpha2 cDNA, we showed that the overexpression of IL13Ralpha2 decreased the response to IL13 but not that to IL4. Our results demonstrate that IL13Ralpha2 acts as a decoy receptor for IL13 and that it may exert a tight regulation of IL13 activity without impairing the IL4 response of the same cell target.