SOD1 in Cerebral Spinal Fluid as a Pharmacodynamic Marker for Antisense Oligonucleotide Therapy

Figure 1. Results of antisense oligonucleotide therapy in rats. Antisense oligonucleotides complementary to SOD1 decreased human SOD1 levels in the brain and cerebral spinal fluid (CSF) of rats. SOD1^G93A rats were dosed with saline (n = 12) or an antisense oligonucleotide complementary to human SOD1 (n = 7) for 1 month, and then the pumps were removed. Two weeks later, CSF and brain samples were harvested. SOD1 messenger RNA (mRNA) was analyzed by use of quantitative polymerase chain reaction (A). SOD1 protein levels in brain (B) and CSF (C) samples were analyzed by use of an enzyme-linked immunosorbent assay and correlated (D). Values are expressed as mean values. Error bars indicate SD. * P < .05, determined by use of the t test. Oligo indicates antisense oligonucleotide.

Figure 2. SOD1 levels in cerebral spinal fluid (CSF) samples obtained from humans. The CSF SOD1 level was determined for individual participants from lumbar punctures completed at the indicated times: 11 participants with amyotrophic lateral sclerosis (each closed symbol or asterisk represents an individual participant with amyotrophic lateral sclerosis), 2 healthy controls (open squares), and 1 control with a neurological disease (open triangle). The average total time participants were observed was 23 months. Although there was substantial variability among individuals' absolute levels of CSF SOD1, there was surprisingly little variability within individuals over time (mean [SD] variation, 7.1% [5.7%]). The maximum variation between any 2 points for a given participant was 21%.

Figure 3. A and B, SOD1 protein levels in cerebral spinal fluid (CSF), which are elevated in participants with amyotrophic lateral sclerosis (ALS) and controls with a neurological disease. In the CSF samples from 88 healthy controls, 89 controls with a neurological disease, and 93 participants with ALS, the SOD1 levels were measured by use of an enzyme-linked immunosorbent assay (A), and total protein levels were measured by use of a bicinchoninic acid assay (B). Values are expressed as mean values. Error bars indicate SE. * P < .05. C, The ratio of SOD1 to total protein was also compared. Of the 89 controls with a neurological disease, 55 received a clinical diagnosis of dementia of the Alzheimer type.

Figure 4. SOD1 protein levels in cerebral spinal fluid (CSF), which do not correlate with disease progression. Of the 93 participants with amyotrophic lateral sclerosis (ALS), 66 had an ALS–Functional Rating Scale (FRS) baseline measurement at the time of the initial CSF collection. The graphical representation of CSF protein vs ALS FRS shows no correlation (R² = 0.001, P = .8) (A). Of the 93 participants with ALS, 62 had additional ALS-FRS measurements performed. The initial CSF SOD1 protein levels did not correlate with the ALS-FRS rate of decline (R² = 0.026, P = .22) (B).