High cerebrospinal fluid levels of interleukin-10 attained by AAV in dogs.

Intrathecal (IT) gene transfer using adeno-associated virus (AAV) may be clinically promising as a treatment for chronic pain if it can produce sufficiently high levels of a transgene product in the cerebrospinal fluid (CSF). Although this strategy was developed in rodents, no studies investigating CSF levels of an analgesic or antiallodynic protein delivered by IT AAV have been performed in large animals. Interleukin-10 (IL-10) is an antiallodynic cytokine for which target therapeutic levels have been established in rats. The present study tested IT AAV8 encoding either human IL-10 (hIL-10) or enhanced green fluorescent protein (EGFP) in a dog model of IT drug delivery. AAV8/hIL-10 at a dose of 3.5 × 10(12) genome copies induced high hIL-10 levels in the CSF, exceeding the target concentration previously found to be antiallodynic in rodents by >1000-fold. AAV8/EGFP targeted the primary sensory and motor neurons and the meninges. hIL-10, a xenogeneic protein in dogs, induced anti-hIL-10 antibodies detectable in the CSF and serum of dogs. The high hIL-10 levels demonstrate the efficacy of AAV for delivery of secreted transgenes into the IT space of large animals, suggesting a strong case for further development toward clinical testing.

Generation and analysis of GluR5(Q636R) kainate receptor mutant mice.

The physiological significance of RNA editing of transcripts that code for kainate-preferring glutamate receptor subunits is unknown, despite the fact that the functional consequences of this molecular modification have been well characterized in cloned receptor subunits. RNA editing of the codon that encodes the glutamine/arginine (Q/R) site in the second membrane domain (MD2) of glutamate receptor 5 (GluR5) and GluR6 kainate receptor subunits produces receptors with reduced calcium permeabilities and single-channel conductances. Approximately 50% of the GluR5 subunit transcripts from adult rat brain are edited at the Q/R site in MD2. To address the role of glutamate receptor mRNA editing in the brain, we have made two strains of mice with mutations at amino acid 636, the Q/R-editing site in GluR5, using embryonic stem cell-mediated transgenesis. GluR5(RloxP/RloxP) mice encode an arginine at the Q/R site of the GluR5 subunit, whereas GluR5(wt(loxP)/wt(loxP)) mice encode a glutamine at this site, similar to wild-type mice. Mutant animals do not exhibit developmental abnormalities, nor do they show deficits in the behavioral paradigms tested in this study. Kainate receptor current densities were reduced by a factor of six in acutely isolated sensory neurons of dorsal root ganglia from GluR5(RloxP/RloxP) mice compared with neurons from wild-type mice. However, the editing mutant mice did not exhibit altered responses to thermal and chemical pain stimuli. Our investigations with the GluR5-editing mutant mice have therefore defined a set of physiological processes in which editing of the GluR5 subunit is unlikely to play an important role.

Fate of the predominant phospholipid component of DepoFoam drug delivery matrix after intrathecal administration of sustained-release encapsulated cytarabine in rats.

The distribution, metabolism, and excretion of dioleoylphosphatidylcholine (DOPC), the predominant phospholipid component of DepoFoam (DF) drug delivery matrix, was determined after lumbar intrathecal injection of double-radiolabeled ((14)CDOPC, (3)H-cytarabine) sustained-release encapsulated cytarabine (DF-cytarabine) in rats prepared with chronic spinal catheters. Radioactivity was quantitated in central nervous system (CNS) and peripheral tissues, cerebrospinal fluid (CSF), blood, urine, and feces at various time points up to 504 h. The distribution of (14)C radiolabel among lipid classes was also determined in selected body fluid samples. Both radiolabels distributed rapidly throughout the neuraxis after injection. Levels of both labels declined in a biphasic manner from CSF and plasma, with an initial rapid decline over the first 96 h, followed by a much slower rate of decline out to 504 h. Greater than 90% of the (3)H (drug) label was estimated to be excreted in urine. In contrast, the data suggest that most of the (14)C (phospholipid) label was expired as (14)CO(2); small percentages of the dose remained incorporated in CNS (7%) and peripheral tissues (8%) or were excreted in urine (6%). Characterization of lipidic (14)C in plasma confirmed metabolism of the parent lipid. The data confirm the sustained-release nature of the DF-cytarabine multivesicular liposomal preparation. Moreover, the results indicate that the DOPC lipid component enters standard catabolic path-ways after breakdown of the DF particles in the intrathecal space. Similar CSF and plasma kinetic profiles of drug and lipid radio-labels support the hypothesis that release of drug is related directly to breakdown of the lipid particles.