No phenotype associated with established lipopolysaccharide model for cerebral palsy

Cerebral palsy (CP) is associated with childhood spasticity, seizures, and paralysis. Oligodendrocyte damage resulting in periventicular leukomalacia (PVL) in the developing brain has been implicated. Animal models of CP have used prenatal hypoxia and infection with histopathology of PVL as the end point. To evaluate whether this histologic end point is associated with a CP phenotype, we reproduced a lipopolysaccharide (LPS) model for PVL, 1 and evaluated developmental, behavioral, and motor outcomes.

On gestational day 15, Fischer 344 rats were intracervically injected with .1 mg/kg LPS (n = 5) or saline (n = 4). After delivery, evaluation for developmental milestones was performed on days 1 to 21 (LPS = 45; control = 30 pups). Males were also tested at 2.5 months using open-field, rotarod, and anxiety tests. On day 21, 2 pups/litter were perfused for immunohistochemistry, and stained with 2 oligodendrocyte antibodies: 2', d'-cyclic nucleotide phosphodiesterase (CNP), and myelin proteolipid protein (PLP) with relative densities of staining assessed using NIH Image software. Statistical analysis included Mann-Whitney U and analysis of variance (ANOVA).

LPS pups demonstrated decreased CNP (P = .04) and PLP (P = .06) staining, replicating the model. There was no difference seen in neonatal weight, righting, negative geotaxis, cliff aversion, rooting, forelimb grasp, audio startle, air righting, eye opening, and activity. Surprisingly, LPS-exposed neonatal rats mastered forelimb placement (P < .01) and surface righting (P = .02) earlier than control rats. There were no differences between adult groups in open field distance traveled (P = .8), open-field locomotion time (P = .6), rotarod (P = .6), or anxiety (P = .7).

Histologic evidence of white matter damage can be replicated using an LPS model for intrauterine inflammation. Significant phenotypic differences consistent with the motor and cognitive damage sequelae of such lesions (ie, CP) were not demonstrated. When evaluating animal models, it is important to assess not only biochemical markers for human disease, but also clinically relevant phenotypes.