ABSTRACT: Missense mutations in the gene for the ubiquitously expressed superoxide dismutase-1 (SOD1) are one of the causes of familial amyotrophic lateral sclerosis (ALS), the most common adult onset motor neuron disease in humans killing selectively large motor neurons. Mice and rats overexpressing mutant SOD1 develop an adult onset neurodegenerative disease with hindlimb-paralysis and subsequent death similar to the human condition. In order to analyze the effects of mutant SOD1 expression onto the most affected cell-type in ALS, a small subpopulation of spinal cord cells, we propose to use laser microdissection to isolate mouse lumbar motor neurons and to assess the changes onto the mRNA expression profile using Affymetrix GeneChips compared to control animals. While two studies applying a genomic approach on the ALS mouse models used the entire spinal cord, contributions of changes to motor neurons were masked by the inflammatory effects of mutant SOD1 and the much larger population of non-motor neuronal cells. What is therefore needed is a cell-type specific expression profile that could reveal dysregulations in the transcriptome of the affected motor neurons. In order to find and analyze disease relevant gene expression profile changes in the mutant vs. the wildtype overexpressing SOD1 mice, we propose to use two different mutant SOD1 lines. Both lines develop the same ALS-like motor neuron disease, however, line 1 (with a G37R mutation) retains the full activity of the SOD1 enzyme while line 2 (with a G85R mutation) has almost no SOD1 enzymatic activity. We believe that by using these two most extreme variants of ALS-inducing mutant SOD1 forms, candidate genes that will be similarly dysregulated in both mutant lines will have great potential to be a disease relevant hit. We will collect from both mutant SOD1 mouse lines lumbar motor neurons from three presymptomatic timepoints, equally distributed during their adult life and before obvious hindlimb-weakness or paralysis signs and compare them to two control lines, negative littermates and wildtype SOD1 overexpressing mice. Furthermore, to determine if their are already mutant SOD1-induced changes since the formation of the affected cells, we will also isolate embryonic spinal cord motor neurons and compare them between mutant SOD1 and wildtype SOD1 overexpressing animals. Both, mice deleted in wildtype SOD1 and mice overexpressing mutant human SOD1 clearly established that the toxic property of mutant SOD1 is based upon a gain-of-toxic function phenomenon rather than a loss-of-function effect, since the SOD1 knock-out mice are overall normal. Furthermore, mice overexpressing wildtype SOD1 are healthy and serve as ideal control animals. Therefore, one of the most important question in ALS is to determine what slowly acting mechanism, or build-up of toxic products is responsible for ultimately killing more than 50-70% of the large lumbar spinal cord motor neurons. We hypothezise that the toxic property of mutant SOD1 must already induce changes onto the transcriptome of the most at-risk cell-type, the large lumbar spinal cord motor neuron, very early on in the presymptomatic phase of the disease course, when the animal is still without any paralysis signs and moving normally around the cage. We will choose 3 timepoints before phenotypic disease-onset that is at 22 weeks for the SOD1G37R line (L42) and at 11 months for the SOD1G85R line (L148). The timepoints are 8, 15 and 18 weeks for SOD1G37R line and 4, 7.5 and 9 months for the SOD1G85R line all compared to age-matched control animals (wildtype SOD1 overexpressing mice (L76) for the SOD1G37R line (high mutant SOD1 levels) and negative littermates for the SOD1G85R line (low mutant SOD1 levels)). Using the Leica lasermicrodissection system on fresh frozen spinal cord sections, we process a single spinal cord sample in 3 days resulting in approximately 4000 ventral horn motor neurons collected from 150 consecutive 20 um sections of the most affected lumbar spinal cord region (L4-L6). To visualize the motor neurons, we have established a fast Nissel-staining protocol that is optimized for speed to preserve the integrity of the RNA. Before lasermicrodissection slides containing 5 spinal cord sections are dessicated for 1 hour and samples are collected directly into RNA-preserving lysis-buffer. Using Stratagene's NanoPrep RNA-extraction columns a yield of around 150 ng of total RNA was obtained (Ribogreen-quantification), that is enough for both the 2-round linear RNA amplification and for future candidate confirmation studies using real-time PCR analysis. We are using the Affymetrix Two-Cycle 3'-Amplification Kit, starting with 100 ng of total RNA and will hybridize on Affymetrix Mouse 430 2.0 GeneChips. Embryonic motor neurons are isolated from e14 mutant SOD1G93A overexpressing rat embryos and compared to wildtype SOD1 overexpressing embryos using a metrazimide gradient followed by a p75-antibody purification in combination with magnetic beads. Cells from one litter of embryos were directly collected into lysis-buffer, yielding at least 200-500 ng of total RNA, enough for one Rat 230 2.0 GeneChip (2-round amplification using 100 ng total RNA). Keywords: time-course
