Modulation of Immune Response by Head Injury - New Insights to Neuroimmune Biology - page 192

performed in rat models. Generally, TNF production seems to be more elevated in
rats than mice, suggesting that genetic differences may influence immune responses
and that the choice of strain may be crucial to the outcome of the experiments [84] .
In clinical studies, TNF has been detected in both CSF and serum of human
patients with severe TBI [55,79-81] . In our patient group, TNF was detected in CSF
only at low concentrations (picogram/ml range), and its pattern did not show an evi-
dent peak, but rather numerous rises above control over 3 weeks [55] . However, in
this patient cohort IL-6 and TGF- were markedly more elevated in CSF compared to
serum, whereas TNF levels were very low and of similar concentration in both fluids.
The contribution of TNF to secondary brain damage is supported by the improvement
in outcome following its neutralization with the administration of soluble binding
proteins, soluble receptors, or immunosuppressive drugs [77,85,86] , whereas injec-
tion of anti-TNF (or anti-IL-6) antibodies had no effect on motor and cognitive out-
come [87] (Ziebell et al., unpublished results).
The role of TNF after TBI became confused as a result of studies employing
TNF or TNF-receptor-gene knockout mice in focal injury paradigms. Deletion of
TNF expression led to a milder deficit in the posttraumatic acute phase but failed to
provide long-term neurological recovery. At 4 weeks, TNF knockout mice also pre-
sented larger lesion volumes compared to wild-type mice [88] . In our study using
TNF/lymphotoxin- double-knockout mice, other than an increased early mortality,
there were no differences in BBB dysfunction, neurological outcome, cell death, or
neutrophil infiltration compared to wild-type TBI controls [72] . Equivalently exacer-
bated tissue and BBB damage in TNF-receptor (p55 and p75) knockouts corroborated
the essential neurotrophic properties of TNF in the pathophysiology of TBI, possi-
bly via activation of the transcription factor NFB, which regulates the expression
of antiapoptotic genes [89] . In addition, earlier we demonstrated that TNF injected
into normal mouse brain reduces constitutive synthesis of the pro-inflammatory
cytokine IL-18 (a member of the IL-1 family), thus reducing brain inflammation [90] .
Altogether, these recent studies suggest that TNF's function may differ in the acute
and the delayed phase after TBI; it seems to act as a potent immune mediator initially,
but as a protective, neurotrophic factor later on, which is required for CNS repair.
First identified as a potent endogenous pyrogen, subsequent research revealed that
IL-1 is a multifunctional cytokine playing a wide range of functions in neuropathologi-
cal diseases, including Alzheimer's, multiple sclerosis, ischemia, depression, and TBI.
Its relevance as a promoter of neurodegeneration predominantly derives from early
studies on brain ischemia [51,91] . IL-1 is rapidly secreted by microglia and its recep-
tors are constitutively expressed by neuronal cells. Among its functions, IL-1 regulates
the secretion of several pro-inflammatory cytokines, chemokines, and reactive oxygen
and nitrogen species, and modulates the activation of matrix metalloproteinases, thus
acting as an initiator of the inflammatory cascade and tissue damage. The duality in
the role of brain inflammation seen with TNF also applies to IL-1. Although microglial
IL-1 is required for astrocyte production of neurotrophins (EGF, CNTF) that promote
regeneration, the deletion of IL-1 in knockout mice failed to achieve proper remyelina-
tion, and these mice showed impaired astrocyte activation that was linked to increased
BBB permeability [92] .