Models of white matter injury: comparison of infectious, hypoxic-ischemic, and excitotoxic insults.

White matter damage (WMD) in preterm neonates is strongly associated with adverse outcome. The etiology of white matter injury is not known but clinical data suggest that ischemia-reperfusion and/or infection-inflammation are important factors. Furthermore, antenatal infection seems to be an important risk factor for brain injury in term infants. In order to explore the pathophysiological mechanisms of WMD and to better understand how infectious agents may affect the vulnerability of the immature brain to injury, numerous novel animal models have been developed over the past decade. WMD can be induced by antenatal or postnatal administration of microbes (E. coli or Gardnerella vaginalis), virus (border disease virus) or bacterial products (lipopolysaccharide, LPS). Alternatively, various hypoperfusion paradigms or administration of excitatory amino acid receptor agonists (excitotoxicity models) can be used. Irrespective of which insult is utilized, the maturational age of the CNS and choice of species seem critical. Generally, lesions with similarity to human WMD, with respect to distribution and morphological characteristics, are easier to induce in gyrencephalic species (rabbits, dogs, cats and sheep) than in rodents. Recently, however, models have been developed in rats (PND 1-7), using either bilateral carotid occlusion or combined hypoxia-ischemia, that produce predominantly white matter lesions. LPS is the infectious agent most often used to produce WMD in immature dogs, cats, or fetal sheep. The mechanism whereby LPS induces brain injury is not completely understood but involves activation of toll-like receptor 4 on immune cells with initiation of a generalized inflammatory response resulting in systemic hypoglycemia, perturbation of coagulation, cerebral hypoperfusion, and activation of inflammatory cells in the CNS. LPS and umbilical cord occlusion both produce WMD with quite similar distribution in 65% gestational sheep. The morphological appearance is different, however, with a more pronounced infiltration of inflammatory cells into the brain and focal microglia/macrophage ("inflammatory WMD") in response to LPS compared to hypoperfusion evoking a more diffuse microglial response usually devoid of cellular infiltrates ("ischemic WMD"). Furthermore, low doses of LPS that by themselves have no adverse effects in 7-day-old rats (maturation corresponding to the near term human fetus), dramatically increase brain injury to a subsequent hypoxic-ischemic challenge, implicating that bacterial products can sensitize the immature CNS. Contrary to this finding, other bacterial agents like lipoteichoic acid were recently shown to induce tolerance of the immature brain suggesting that the innate immune system may respond differently to various ligands, which needs to be further explored. Copyright 2002 Wiley-Liss, Inc.