Persistent poliovirus infection(F. Colbère-Garapin)
Enteroviruses can persist in immunodeficient individuals, in particular in those with agammaglobulinemia. Infection may remain undetected for years, until a chronic disease eventually develops. Indeed, recent studies have shown that some immunodeficient individuals can harbor and excrete poliovirus (PV) for more than 10 years.
PV is the etiologic agent of paralytic poliomyelitis. It is among the best characterized enteroviruses, and the structure of the capsid, the viral genome and the multiplication cycle have been extensively studied. It is considered to be the proptotype enterovirus. We have shown that PV can establish persistent infections in human cells of neuronal origin in culture, and have also developed several models to elucidate the molecular mechanisms of persistent PV infections.
During persistent PV infection of neuroblastoma IMR-32 cells, mutant viruses (PVpi) are selected. Some PVpi mutations affect capsid residues located in regions involved in the interactions with the virus receptor (poliovirus receptor: PVR, or CD155). This led us to study PVR in persistently infected cells. PVR is a protein belonging to the immunoglobulin superfamily; it has three extracellular domains, a transmembrane region and a cytoplasmic region. Only the first extracellular domain directly interacts with PV. In mRNAs of persistently PV-infected IMR-32 cells, we repeatedly identified mutations all localized in the first domain of PVR (N. Pavio et coll., Virol., 2000, 274, 331-342). These mutations have been introduced into the cloned PVR cDNA, and the mutated PVRs have been introduced into PVR-negative cells by transfection. The PVR mutations confer on cells a partial resistance to wild-type virus-induced lysis. The adsorption of virus onto mutant receptors is not modified, but there is a decrease in the efficiency of conformational modifications of the capsid induced by the PVR and believed to be necessary for viral penetration and uncoating (collaboration with B. Blondel & T. Couderc's group). This suggests that the efficiency of virus uncoating induced by the PVRs is reduced by the mutations.
This work has been continued with PV mutants having point mutations from PVpi (selected in neuroblastoma and in human fetal brain cells). After infection, murine cells expressing mutant forms of PVR are more resistant to lysis than cells expressing PVR of IMR-32 cells, independent of the virus strain. The kinetics of adsorption of the mutant viruses onto the PVR of IMR-32 cells are similar to those of the wild-type virus, but the conformational modifications induced by the PVR are altered by the PV mutations. Thus, persistent PV infections in neuroblastoma cells appear to be maintained by a molecular mechanism involving modifications of virus-receptor interactions as a result of co-evolution of the virus and cell.
PVpi selected in human neuronal cells are able to establish persistent infections in nonneural cells, whereas their parental viral strains are not. We have identified several viral determinants involved in this phenotype. A single determinant is in some cases sufficient to confer to a lytic virus the capacity to establish persistent infections in nonneural HEp-2 cells, but the association of several determinants has a synergistic effect. These determinants profoundly affect the early steps of the virus cycle, including cell binding and the receptor-mediated conformational changes of the capsid. Some persistent mutants appear to undergo a novel capsid transition when in contact with the PVR. An abnormal stability of particular PV uncoating intermediates may delay the entry of the viral genome into the cell. We are currently studying the early steps of infection with purified soluble PVR molecules (collaboration with A. Nomoto & M. Arita (Tokyo)).
J. Martin (N.I.B.S.C., United Kingdom) has characterized PV strains isolated from a hypogammaglobulinemic individual, who remained asymptomatic while excreting PV for almost two years (J. Martin, et coll. J. Virol. 2000, 74, 3001-3010). These strains were derived from the type 3 Sabin 3 strain. Some of the mutations are identical, or very close to those that we detected in the genome of PVpi persisting in human neuroblastoma cells (G. Duncan et coll., Virol., 1998, 241, 14-29). PV persistence has never been studied in an in vitro model of human intestinal cells. To elucidate how PV persists in humans, we are developing a model in collaboration with Javier Martin, using the isolates from the hypogammaglobulinemic patient and human intestinal cells. This should allow us to evaluate the respective importance of viral and cellular determinants in the molecular mechanisms of this persistent infection.
Poliovirus as a model for studying virus-nerve cell interactions : apoptosis and persistence (B. Blondel and T. Couderc).
Neurotropic viruses can persist in the central nervous system following the acute phase of infection and induce new pathologies several years after the initial infection. Poliovirus, a member of the Picornaviridae family, is the aetiological agent of poliomyelitis. It is currently one of the best-characterized neurotropic viruses. Patients having recovered from acute poliomyelitis developed after several decades of clinical stability a new disease, called post-polio syndrome, characterized notably by slowly progressive muscle weakness and atrophy. One hypothesis to explain this syndrome could be poliovirus persistence in the central nervous system, possibly associated with an immunopathological process.
We have previously developed a mouse model susceptible to poliovirus infection and we have shown that poliovirus can persist in the central nervous system after the onset of paralysis throughout the life of animals. We have also shown that the poliovirus persistence could be due, at least in part, to an inhibition of viral genome synthesis in the central nervous system. During the acute phase of poliomyelitis, we have demonstrated that poliovirus kills motoneurons by an apoptotic process.
We have recently developed a model of mixed mouse primary nerve cell cultures to study the molecular mechanisms of poliovirus-induced apoptosis in nerve cells. We are currently investigating the role of interactions of poliovirus with its cellular receptor (CD155) and that of transduction signal in the induction of apoptosis versus persistence. In particular, we are analyzing caspases activation and mitochondrial dysfunctions as well as NF-k
B activation following poliovirus/CD155 interaction.
Finally, mice surviving paralytic poliomyelitis represent a relevant animal model to study processes leading to regeneration of paralyzed muscle following virus-induced motoneuron death.
Infection with rabies virus (M. Lafon)
Rabies virus is an enveloped RNA virus of the rhabdoviridae group. Rabies virus is inoculated by bite of a rabid animal. It enters the nervous system either by motor neuron, through the neuromuscular junction, or by sensory nerve, through nerves spindles. Then it travels from one neuron to the next one along the spinal cord up to brain, and salivary glands. It modifies the animal behavior, (furious state) by triggering dysfunction of gaba-cholinergic neurons. Then, virus particles are excreted into the saliva of the furious animal and thus transmitted to another host.
Rabies virus has been mouse adapted.. Several mutants of different pathogenicity have been selected and are well characterized.
In the laboratory, we use rabies virus as a model of viral encephalitis to study the interactions of rabies virus with the nervous system.
Firstly at a molecular level by demonstrating that NCAM is a main receptor used by rabies virus to invade the nervous system and by analysing the virally-induced apoptosis
Secondly, by studying the mechanisms developed by pathogenic strains of rabies virus to subvert the host surveillance (strengthening of the immune privilege of the nervous system NS and induction of a peripheral immuno-suppression).
(1) The Neural Cell Adhesion Molecule (NCAM) is a Receptor for Rabies Virus. Implication for tropism and pathogenicity.
We found that laboratory cell lines susceptible to rabies virus infection express the Neural Cell Adhesion Molecule NCAM (CD56) on their surface, whereas resistant cells do not, supporting the idea that NCAM could be a rabies virus receptor. We observed that 1) incubation with rabies virus decreases the surface expression of NCAM, 2) treatment of susceptible cells with heparan sulfate, a ligand for NCAM, or with NCAM-antibodies significantly reduces the rabies virus infection and 3) pre-incubation of rabies virus inoculum with soluble NCAM protein as a receptor decoy drastically neutralizes the capacity of rabies virus to infect susceptible cells. In addition, we demonstrated that transfection of resistant L fibroblasts with the NCAM gene induces rabies virus susceptibility whereas absence of NCAM in the primary cortical cell cultures prepared from NCAM deficient mice reduces the rabies virus infection and virus production. This provides evidence that NCAM is an in vitro receptor for the rabies virus.
NCAM is a polymorph protein with variable size of cytoplasm tail and level of sialilation (PSA). By transfection of different NCAM isoforms, polysialic acid removal and the use of PSA-KO mice, we identify that rabies virus utilizes an isoform combination mainly present on the neurons of adult nervous system. This is consistent with the tropism of rabies virus for neurons.
Moreover, the in vivo relevance for the use of NCAM as receptor was demonstrated by the infection of NCAM deficient mice, in which rabies mortality was delayed and brain invasion by rabies virus was drastically restricted. Our results showed that NCAM, that is mainly expressed in the adult nervous system, and at the neuro-muscular junctions, plays an important role in rabies infection. However since mice lacking NCAM died, it cannot be excluded that receptors other than NCAM are utilized, in particular the nicotinic acetylcholine receptor (nAchR). The pentameric a7nAchR brain receptor, that is susceptible to abungarotoxin, could be a second-order receptor used by rabies to travel into the nervous system. This would be another example of the use by viruses of more than one receptor to gain entry into the host. By comparing progression of infection in the brain and by using transfected cells we showed that in contrast to the neurotropic strains of virus, the non neurotropic ones cannot use NCAM, indicating that NCAM plays an important role in neurotropism and pathogenicity. These findings bring strong support to design novel anti-viral strategies using ligand molecules to mask viral receptors on the cell surface.
(2) Pathogenicity of rabies virus correlates with the conservation of the CNS immune privilege
We compared the components of the immune response in the CNS after an infection either by an acute rabies virus infection or an abortive one. Both pathogenic viruses strain and those with attenuated pathogenicity are neurotropic and replicate in neurons equally. However the pathogenic virus strain induces a fatal acute encephalitis, whereas the abortive infection results in a non fatal disease characterized by transient infection and irreversible paralytic sequels of the inoculated limbs.
In the course of acute rabies encephalitis, we observed that infection triggers chemokines and inflammatory cytokine production and a transient migration of activated lymphocytes in the CNS.. However, these lymphocytes, which specificity is unknown, do not play any role in the control of the infection since severity of rabies is similar in nude and immuno-competent mice (Camelo et al, 2001a). T cells migrate into the CNS. However this migration that peaks 6 days after infection is transitory since migratory T cells have disappeared by day 9. Antibodies are not detectable in the CNS.
In contrast, in the course of abortive rabies, our data indicate that lymphocytes are essential for protection, since nude mice develop a fatal encephalitic form of rabies (Galelli et al., 2000). We observed that infection triggers chemokines and inflammatory production. CD4 and CD8 T cells migrate into the CNS. CD8 T cells are still observed 14 days after the infection whereas CD4 declined by day 10. Abortive rabies is characterized by paralytic sequels. Mice deleted of CD8 T cells were not paralyzed, indicating that in this immuno pathology was under the control of migratory CD8 T cells.
Large quantities of antibodies and B cells can be observed in the CNS. Others indicate that antibodies and CD4+ are the key factors for survival (Hooper et al, 1998).
A striking difference between pathogenic and non-pathogenic rabies virus strains is that pathogenic strains do not destroy infected neurons, whereas abortive virus strain do. Virus-induced apoptosis is both a direct and a bystander mechanisms since both infected neurons and non-infected neuronal cell encounter apoptosis. Nature of the noninfected neurons that encounter apoptosis during rabies virus CNS infection is unknown; We found in retina of rabies virus infected animals that rabies virus infection is sequestrated in the ganglion cells. Nevertheless, the infection triggers a bystander apoptosis in the photoreceptors two layers above. Identification of mechanisms involved requires further attention.
It seems therefore that rabies virus pathogenicity is linked to the absence of neuronal apoptosis. This is uncommon, since it is well-accepted that, pathogenicity is associated to virus-mediated destruction of nervous tissue. It seems that in the case of CNS infection by rabies virus , the induction of apoptosis is a key factor to trigger an immune response which limits the spread of the infection.
Our hypothesis is that during an acute encephalitis, the immune privilege is drastically maintained because there is no antigen to be presented to the immune system. Pathogenic strains do not trigger apoptosis. There is no release of apoptotic bodies we found to be strong immunogens (see paragraph 3 of this report, page 6). We propose that weakness of the immune response results of the lack of antigen to be sampled by antigen presenting cells (APC) and presented to the immune peripheral system. (see scheme below)
In contrast, in a non-pathogenic infection, neurons are destroyed, viral antigen embedded in apoptotic bodies are picked up by professional APC (such as dendritic cells of the choroids plexus). These cells migrate to the cervical lymph nodes. A primary immune response directed against the viral antigens can take place. Antibodies triggered with the help of CD4 T cells, clear the infection. However, a dual role is assigned for the CD8 T cells : they participate in the CNS clearance but they induce neuron apoptosis (bystander mechanism) and thus initiate paralytic sequels. Mice survive despite irreversible paralytic sequels.
Collectively these data indicate that protection against apoptosis could be a key factor of rabies virus pathogenicity. In contrast, clearance of CNS infection requires apoptosis of infected neurons in part because it interrupts the axonal flow of virus transmission, in part because it triggers an immune response.
(3) Apoptotic bodies from virus infected lymphocytes are powerful immunogens
Rabies can be prevented in humans by a post exposure vaccination. Antibodies produced with the help of CD4 T cells are the main component of the protective response induced by these vaccines. After dentritic cells engulfed apoptotic cells resulting from the destruction of tumor cells and presentation of self antigens, they induce tolerance (Huang et al, 2000, and Sauter et al, 2000,). In contrast , when dentritic cells, phagocyte viral infected apoptotic cells, they cross present viral antigens to CD8+T cells and induce a cytotoxic response (Albert et al,, 1998). We tested the possibility that apoptotic bodies could also stimulate antibodies production. We compared the production of rabies virus specific antibodies and neutralizing antibodies in mice injected with rabies virus infected cells. Mice immunized with apoptotic infected cells produce ten times more rabies antibodies than non fragmentized cells. This indicates that apoptotic bodies engulfed by APC cells such as immature dendritic cells, are an important source of antigen efficiently presented to the immune system. Altogether these data indicate that induction of apoptosis and uptake of apoptotic bodies by migrating DC could be important factors for the efficacy of post exposure rabies vaccination.
We tried to elucidate why live attenuated rabies virus strain used as oral vaccines to eradicate rabies from the European wildlife (mainly foxes) were so efficient. The better immunogenicity of live rabies vaccines compared to killed vaccines could result of the virus replication in peripheral tissues such as macrophages or lymphocytes. These vaccines are prepared with laboratory strains which lost their capacity to replicate in the nervous system (ERA, SAD..). We demonstrated these strains infect lymphocytes which then died by apoptosis. Since lymphocytes are corner stones of the immune response, it was rationale to postulate the existence of an immunopotent mechanism which enhances the immune response and counteracts the lymphocyte lost.
We presented evidence that immuno-potentiation mechanisms result from the following factors :1) rabies virus infects only a fraction of lymphocyte population, 2) apoptosis occurs too late in the course of the virus cycle to hinder virus production 3) apoptotic bodies are powerful inducers of immune response, 4) apoptosis releases viral NC which is a powerful adjuvant for any associated vaccination and 5) finally because INF- is released in the course of lymphocytes infection.
(4) Immune unresponsiveness
Previous reports suggest that rabies virus can subvert the host immune response.. In contrast to infections such as measles virus or HIV, in which immunosuppression results of the destruction of immune cell effectors, rabies virus induced-immunosuppression is a consequence of the CNS infection, because rabies virus does not replicate outside the CNS. Thus, rabies virus CNS infection is a good model to analyze how stress (infectious stress in particular) controls the immune response in the periphery.
Fatal encephalitis induced by pathogenic strains of rabies virus, such as CVS, or street rabies virus, is accompanied by an immuno-depression characterized by the impairment of the lymphocyte response marked by a loss of cellular mediated immunity (Wiktor et al, 1977). This is concomitant with the unresponsiveness of spleen cell to ConA re-stimulation in vitro.The decline in lympho-proliferative response of splenocytes to an in vitro ConA stimulation but LPS from the 6th day of infection is associated to a decrease in the number of Th1 (IL-2, IFN-g and TNF a but not of Th2 (IL-4) secreting splenocytes. No quantitative modifications of the different spleen populations (CD4, CD8, B cells, NK..) were observed. The decrease of immune responsiveness depends on the pathogenicity of the strain. Abortive strains of rabies virus such as the Pasteur Virus strain, PV, do not induce immunosuppression suggesting that a certain threshold of dysfunction and/or inflammation of the nervous system is required. In the contrary of the lymphoid depletion that also characterizes rabies virus infection, the immune suppression is not under the control of the HPA axis, since adrenalectomy does not modify the CTL reactivity (Wiktor et al, 1977). Besides the contribution of HPA axis in the control of immune response by the CNS, there is now lines of evidence that the control of nervous on immune systems can also be exerted through a direct local delivery of neurotransmitters to peripheral immune organs by efferent nerves. Nerves deliver locally neurotransmitters, catecholamines (nor epinephrine in particular) to the immune cells bearing the appropriate receptors, leading them to secrete cytokines such as IL-6, TNF-a or Il-10 that can perturb the immune response (Straub et al, 1998). Characteristics of the rabies induced un responsiveness lead us to suspect involvement of NE in this control. Exact mechanisms of NE control in rabies remain to be elucidated.
Involvement of TNF-a in the rabies immune un responsiveness was tested by using mice lacking the TNF-a receptor (p55). By comparing ConA responsiveness in CVS infected mice lacking the TNF-a receptor (p55) and in their normal counterparts, it appears that immune unresponsiveness is under the control of the p55 receptor and is not linked to the loss of MHC class II expression nor the altered CD4/CD8 ratio in spleen which are observed during rabies virus CNS infection. Immune un responsiveness is associated with an upregulation of CD69 and CD25 on the surface of spleen CD3 cells. This could be indicative of an involvement of regulatory T cells (former suppressor Tcells). Alternatively immune unresponsiveness induced by rabies virus infection could be the result of a default in the T cells activation in a context of poor antigen presentation leading to the unresponsiveness of T cells and the reduction of cytokines secreting cells . Neuronal immunosuppression induced by the viral stress of rabies infection can strengthen the subversion of the immune system by this CNS infection. It could represent an additional factor in the global unresponsiveness of the host to rabies.
Myelin-forming cells in development and disease (Monique Dubois-Dalcq)
A Signals promoting oligodendrocyte development from neural precursors.
(1)Sonic hedgehog. Murray, Rottkamp , Calaroaet al, in preparation
The potential to generate OP from neural stem cells (NSCs) exists throughout the developing CNS. Yet, they emerge only in restricted regions of the developing spinal cord in response to the a morphogen Sonic hedgehog (Shh). As the territory of Shh expression
extends to the ventral telencephalic region in rodents, we have investigated whether it can trigger oligodendrocyte emergence in the developing forebrain. Shh and its receptor Patched were detected by RT/PCR in subventricular zone (SVZ) tissues of newborn rats
at the time of intense gliogenesis. In contrast Shh was not detected in E17 cortex where no oligodendrocytes had developed yet. We then tested the ability of recombinant Shh to induce OP in E17 cortical explants. After 3 days, OP emerged one day earlier in the
presence of Shh than in controls. Two days later, we observed an Shh dose dependant increase in the percentage of OP with a maximum of a ten fold increase at 10 µg/ml. Using cocultures of cortical precursors with quail cells expressing Shh (courtesy of Dr Duprez), we observed a doubling of the mitotic index in oligodendrocyte progenitors and up to a 9 fold increase in cells expressing oligodendrocyte specific glycolipids. These data suggest that, as in the embryonic spinal cord, Shh can signal cortical precursors to become oligodendrocytes in the telencephalic region.
(2)Neuregulins (NRG) Calaora et al, in press in J Neuroscience
Neuregulin 1 (Nrg-1) isoforms have been shown to influence the emergence and growth of oligodendrocytes, the CNS myelin-forming cells. We have investigated how Nrg-1 signaling of ErbB receptors specifically controls the early stages of oligodendrocyte generation from multipotential neural precursors (NP). We show here that embryonic striatal NP express multiple Nrg-1 transcripts and proteins, as well as their specific receptors, ErbB2 and ErbB4, but not ErbB3. The major isoform synthesized by striatal NP is a transmembrane Type III isoform called Cystein-rich domain Nrg-1. To examine the biological effect of Nrg-1, we added soluble ErbB3 (sErbB3) to growing neurospheres. This inhibitor of Nrg-1 bioactivity decreased NP mitosis and increased their apoptosis, resulting in a significant reduction in neurosphere size and number. When NP were induced to migrate and differentiate by adhesion of neurospheres to the substratum, the level of type III isoforms detected by RT/PCR and Western blot decreased in parallel with a reduction in Nrg-1 fluorescence intensity in differentiating astrocytes, neurons and oligodendrocytes. Pretreatment of growing neurospheres with sErbB3 induced a three fold increase in the proportion of oligodendrocytes generated from NP migrating out of the neurosphere. This effect was not observed with an unrelated soluble receptor. Addition of sErbB3 during NP growth and differentiation enhanced oligodendrocyte maturation as shown by expression of galactocerebroside and myelin basic protein. We propose that both Type III Nrg-1 signaling and soluble ErbB receptors modulate oligodendrocyte development from NP.
B Effects of maintained expression of PSA-NCAM in neural precursors
(1) Retroviral Gene Transfer in Mouse Neural Precursors Franceschini et al, in press in JNR
Gene transfer into neural precursors is a powerful approach to study the function of specific gene products during nervous system development. Here, we describe a retrovirus-based methodology to transduce foreign genes into mouse neural precursors. We used a high titer bicistronic retroviral vector which encodes a marker gene, placental alkaline phosphatase (plap) and a selection gene, neomycin phosphotransferase II (neoR), under the translational control of two retroviral internal ribosome entry segments. Transduction efficiency even without selection was up to 95 % for multipotential neurospheres derived from embryonic striata and grown with basic fibroblast growth factor 2. Expression of plap and neoR was sustained with time in culture and upon differentiation into neurons, astrocytes and oligodendrocytes, as shown by double immunofluorescence labeling with cell type-specific markers, Western blotting and neomycin resistance. However, levels of plap were decreased in differentiated oligodendrocytes. Transduction with the same vector of neonatal oligodendrocyte precursors grown in oligospheres consistently resulted in a lower proportion of plap-immunoreactive cells and enhanced cell death in the absence of neomycin. Yet, plap expression was maintained in some differentiated oligodendrocytes expressing galactocerebroside or myelin basic protein. Since neurospheres can be easily expanded in vitro and since factors enabling their differentiation into the three main central nervous system cell types are being elucidated, this methodology could be used in the future to produce large number of transduced differentiated neural cells.
(2) Genetic manipulation of neural stem cells to study the role of polysialylated neural cell adhesion molecule (psa-ncam) in oligodendrocyte development. I.A. Franceschini1 et al, coll with Burnham Institute, La Jolla, CA, USA. Abstract for the meeting of the American Neuroscience Society, San Diego, CA, November 2001
Recently, we described an efficient retrovirus-based method to transduce foreign genes into mouse neural precursors (Franceschini et al., JNR, in press). We used this approach to transfer the polysialyltransferase STX cDNA into multipotent precursors from embryonic striata, expanded as neurospheres with EGF and/or FGF-2. STX synthesizes the polysialyl groups on NCAM implicated in neural plasticity during development and regeneration. Non transduced neurospheres were weakly immunoreactive for PSA-NCAM as were neurospheres transduced with vectors encoding green fluorescent protein or alkaline phosphatase. In contrast, neurospheres transduced with STX were strongly immunoreactive for PSA-NCAM, a characteristic maintained over numerous passages. After neurosphere adhesion, the three CNS lineages emerged normally from these genetically ingeneered precursors, as identified with GFAP (astrocytes), beta3 tubulin (neurons) and O4 (oligodendrocytes) antibodies. However the O4-positive progeny of the STX-transduced neurospheres appeared different: the proportion of cells with complex processes and/or O4 positive sheaths was reduced and fewer cells expressed myelin basic protein. Immature looking O4 positive cells with fewer processes were more numerous. When neurospheres were adhered to matrigel, numerous cells migrated out in chains and STX-transduced neurospheres developed more chains than the other neurospheres. These culture models should be useful to unravel the role of PSA-NCAM in migration and differentiation of oligodendrocyte precursors.
C Molecular Pathogenesis of a genetic disease affecting CNS myelin: X-linked adrenoleukodystrophy.(coll with INSERM, St Vincent de Paul/Necker,).
Aubourg and Dubois-Dalcq, Glia, 29, 186-190, 2000 ; Feigenbaum et al., Neurobiology of Disease, 7, 600-612, 2000.
Mutations in the ALD gene coding for an ABC transporter in the peroxisomal membrane of glial cells can lead to a demyelinating disease with a rapid fatal outcome in boys. ALD mutations tend to cluster in this transporter functional regions without correlating with a particular clinical course. Is the increase in very long chain fatty acids in blood and brain the disease trigger? The ALD protein is expressed in astrocytes, microglia and oligodendrocytes but not in neurons. How does it cause myelin, once formed, to degenerate and bring inflammation into this genetic disease? Would apoptosis of myelin-forming cells be a primary event? To investigate whether apoptotic processes take place in the ALD brain, we examined autopsy specimens from telencephalic and brainstem regions of four patients and compared these tissues to those of three controls. Two ALD patients showed internucleosomal DNA fragmentation, suggesting a high number of apoptotic cells in affected white matter regions. By this method, none of the controls showed apoptosis in either brain white matter or brainstem. In three ALD patients, in situ detection of DNA breaks by the TUNEL method revealed stained nuclei with chromatin alterations in areas of demyelination. Identification of the cells with TUNEL positive nuclei showed that up to 47% of apoptotic nuclei pertained to oligodendrocytes. Caspase-3 immunostaining was also detected, indicating that programmed cell death was taking place in these cells. We conclude that apoptosis of oligodendrocytes may account, at least in part, for the demyelinating process in the ALD brain.
D Alpha chemokine in the CNS: CXC-R4 signalling by Stromal Cell-derived factor 1 and it function in the rodent brain , in particular neural precursors
Lazarini et al European J Neuroscience, 12, 1-9, 2000; To Nam Tham et al European J Neuroscience, 13, 845-56, 2001
CXCR4 is the Gi protein-linked seven-transmembrane receptor for the alpha chemokine Stromal Cell-Derived Factor 1 (SDF-1). SDF-1 is an alpha-chemokine that stimulates migration of hematopoietic progenitor cells and development of the immune system. Its receptor, CXC-R4, is highly conserved between human and rodent. CXCR4 is also a coreceptor for entry of human immunodeficiency virus (HIV) in T cells and is expressed in the CNS. To investigate how these CXCR4 ligands influence CNS development and/or function, we have examined the expression and signalling of this chemokine receptor in rat neurons and astrocytes in vitro. CXCR4 is expressed in both cell types and in E15 brain neuronal progenitors. CXCR4 signalling by SDF-1 of these progenitors induced activation of extracellular signal regulated kinases (ERKs) and a dose-dependent chemotactic response. In differentiated neurons, both SDF-1 and the glycoprotein of HIV, gp120, triggered activation of ERKs and cjun amino-terminal kinase (JNK1). These effects were significantly inhibited by Pertussis toxin, which uncouples Gi proteins, and by the bicyclam AMD3100, a highly selective CXCR4 antagonist. Rat astrocytes also responded to SDF-1 signalling by phosphorylation of JNK1 and ERKs but, in contrast to cortical neurons, no kinase activation was induced by gp120. Thus neurons and astrocytes can respond differently to signalling by SDF-1 and/or gp120. As SDF-1 triggers directed migration of neuronal progenitors, this alpha chemokine may play a role in cortex development. In differentiated neurons, both natural and viral ligands of CXCR4 activate specific kinases and may therefore influence neuronal function.
In support of this idea is the developmental pattern of Expression of SDF-1 in the rat central nervous system. SDF-1 is abundantly and selectively expressed in the developing and mature CNS. At embryonic day 15, SDF-1 transcripts were detected in the germinal periventricular zone and in the deep layer of the forming cerebral cortex. At birth, granule cells in the cerebellum and glial cells of the olfactory bulb outer layer showed an SDF-1 in situ hybridization signal that decreased progressively within the next two weeks. In other regions such as cortex, thalamus and hippocampus, SDF-1 transcripts detected at birth progressively increased in abundance during the postnatal period. SDF-1 protein was identified by immunoblot and/or immunocytochemistry in most brain regions where these transcripts were detected. SDF-1 was selectively localized in some thalamic nuclei and neurons of the fifth cortical layer as well as in pontine and brainstem nuclei which relay the nociceptive response. The presence of SDF-1 transcripts in cerebellar granule cells was correlated with their migration from the external to the inner granular layers with disappearance of the signal when migration was completed. In contrast, SDF1 mRNA signal increased during formation of the hippocampal dentate gyrus and stayed high in this region throughout life. The selective and regulated expression of SDF-1 in these regions suggests a role in precursor migration, neurogenesis and, possibly, synaptogenesis. Thus, this alpha chemokine may be as essential to nervous system function as it is to the immune system.
Study of cellular and molecular mechanisms of diseases associated to connexins mutations (Roberto Bruzzone)
Most cells communicate with their immediate neighbors through the exchange of cytosolic molecules such as ions, second messengers and small metabolites. This activity is made possible by clusters of intercellular channels called gap junctions, that connect adjacent cells. The molecular architecture of intercellular channels consists of two channels, called connexons, which interact to span the plasma membrane of two adjacent cells and directly join the cytoplasm of one cell to another. Connexons are made of structural proteins named connexins, that compose a multigene family whose members are distinguished according to their predicted molecular mass in kDa. Connexin channels participate in the regulation of signaling between developing and differentiated cell types. The recent discovery of human genetic diseases associated with mutations in six connexin genes and the study of knockout mice have validated the view that this form of intercellular signaling fulfills a crucial role in coordinating several aspects of tissue homeostasis. Understanding in detail how these channels gate offers the potential to develop specific drugs to deal with connexin-based disorders. To address this issue, we propose to take advantage of the naturally occurring connexin mutations to gain insight into the regulatory mechanisms of channel gating and further develop a related project on the specific role of connexins in the central nervous system (CNS), a largely unexplored area. Three questions are being specifically investigated: (1) What are the cellular and molecular basis of connexin diseases? (2) What are the functional properties of the novel subgroup of neuronal connexins? (3) What is the role of connexins in CNS development and differentiation?
Connexins and human diseases
Mutations in six connexin genes have been linked with five different pathologies. Mutations in Cx32 cause X-linked Charcot-Marie-Tooth disease (CMTX), the second most common inherited demyelinating neuropathy of the peripheral nervous system, while mutations in either Cx46 or Cx50 cause autosomal dominant cataracts . Three connexins, Cx26, Cx30, Cx31, have been linked to different forms of dominant and recessive nonsyndromic deafness, one of the most prevalent inherited sensory disorders. Cx26 mutations can also underlie syndromic forms of hearing loss that are associated with palmoplantar keratoderma, whereas Cx31 has been linked to erythrokeratodermia variabilis.
Connexins in the retina
Gap junctions play a key role in the functional organization of the vertebrate retina, as virtually every cell type is coupled to its neighbors by these intercellular channels. Moreover, a broad range of experimental studies has shown that the gap junctions connecting different types of retinal cells are selectively permeable to small tracers, exhibit unique physiological and pharmacological properties, and are differentially gated. Curiously, none of the distributions or functional properties of the known connexins studied in vitro is in good agreement with the distinct pharmacological properties of different gap junctional pathways in the vertebrate retina. The recent identification and cloning of members of the neuronal connexin subgroup make available, for the first time, participants in gap junction formation between various retinal neurons for study and reconstitution in experimental systems.
Connexins in the developing nervous system
It has been proposed that connexin channels are needed for the electrical and biochemical coordination of many processes, including cell migration, differentiation and formation of synaptic circuits. As development proceeds, connexins are expressed in complex and overlapping patterns whose significance is still unclear. One potential consequence of this program is the establishment of boundaries or gradients of communication between contacting cells, as formation of gap junction channels is a selective process that is also dependent on the presence of compatible connexins in adjacent cells. Gradients or boundaries of communication would result in distinct signals being sensed by neighboring cells, leading to the activation of different gene expression programs within a group of cells. Thus far, it is not known whether connexin-based communication pathways plays are implicated during specific steps of neurogenesis and gliogenesis. The regulation of the spatial and temporal expression of connexins during development makes it attractive to elucidate their role in the series of events leading to the emergence of differentiated cells from a pool of precursors.