Expression of mRNAs in mammalian spinal cord at times when regenerating can and cannot occur after injury

The aim of the present experiments was to analyze molecular mechanisms responsible for enhancing and preventing regeneration in the mammalian central nervous system (CNS). Unlike the situation in adult mammals, the CNS of the neonatal opossum shows successful regeneration after injury. The transition point has been shown to occur in the early days of postnatal life. In the newborn opossum the potential for repair becomes drastically diminished in the cervical spinal cords of animals aged 12 days or more. Owing to a rostrocaudal gradient in development, animals as old as 17 days still show regeneration in less mature lumbar segments of the spinal cord. The differences between opossum nervous systems at different stages of development that can and cannot regenerate were used to determine what changes in gene expression occur during the critical period for regeneration. As a first step methods were developed for identifying those mRNAs that are differentially expressed in regenerating and non-regenerating opossum spinal cord. To analyze the differences between 9 and 13 days we used suppression subtractive hybridization. This method, based on PCR, allows the identification of transcripts expressed in one population but not in the other. In addition, the cDNAs that are differentially expressed are selectively amplified. Both forward and backward subtractions were performed between cervical spinal cord mRNA from 9-day and 13-day animals. Comparable subtractions were made with cervical (unable to regenerate) and lumbar (still able to regenerate) spinal cord at 13 days. Additional subtractions were made between uninjured and lesioned spinal cords. The cloning of cDNAs from the 9 - 13 and 13 - 9-day subtractions resulted in about 3000 clones. 88 clones were picked at random and sequenced. About 50% represented novel genes whose function is yet unknown. A large number of potentially interesting genes were identified. They included membrane proteins, signal regulatory proteins, transcription factors, enzymes, cytoskeletal proteins. Examples of upregulation in regenerating tissue are: rig protein, known to be expressed in regenerating liver, alpha-spectrin, tenascin-C, protein kinase C inhibitor. Among genes more abundant in non-regenerating tissue were found: BTG-1 antiproliferative factor, calmodulin 2, Mo25 (calcium binding), profilin 2 (actin binding). Sequencing also revealed the presence of ubiquitously expressed genes like ribosomal proteins. All identified genes showed high similarity to human, rat and mouse genes. From a study of this type many genes are amplified that have nothing to do with regeneration. They are developmentally regulated genes, and high-abundance clones that escaped the process of subtraction. To separate non-relevant genes from those involved in regeneration we designed experiments based on cross-hybridization between different subtractions. All the clones from the subtractive libraries were arrayed onto nylon membranes and hybridized with radioactively labeled probes from other subtractions. This process reduced the number of potentially involved genes by a factor of about ten. Sequencing identified individual positive genes. These include among others reelin, laminin receptor, TATA-box binding protein, MAPl b, all of which were more abundant at 9 days. At 13 days the genes selectively expressed included P AX-6, calmodulin, semaphorin receptor, GABA(A) receptorassociated protein. Novel genes were represented in both categories. Additionally, membrane hybridization of the positive clones with radiolabeled first strand cDNA prepared from Tester and Driver was performed. This step removed highabundance transcripts. In contrast with other studies focusing on one gene or protein, we have carried out an approach based on broad screenings that can lead to unbiased identification of potentially interesting genes in the spinal cord at different ages. The distribution and quantitation of candidates have been confirmed by Northern blotting, RNAse protection, and In situ hybridization. Together with antibody studies to localize the proteins of interest, these are essential steps required for controlling that the genes detected are not artifacts of PCR amplification, that they are present in the original tissue, as well as their precise location and abundance. The next steps planned include the introduction of candidate genes into the lesioned opossum spinal cord and testing their effect on regeneration