Our observations suggest an important role for the cytokine IL-16 in regulation of CD4+ T cell infiltration, of severity and frequency of relapses, and of subsequent demyelination and axonal damage, in a mouse model of EAE . These observations suggest that IL-16, which serves as a specific ligand for the CD4 co-receptor, may have a similar role or roles in regulation of MS. In other organs, increased levels and processing of IL-16 is associated with the pathogenesis of delayed type hypersensitivity  and autoimmunity, such as in atopic dermatitis , with autoimmune rheumatoid arthritis and lupus [25, 26]. However, the role of IL-16 regulation in MS lesions is largely unknown. With the exception of one report, which shows IL-16 immunoreactivity within lesions of a single case of MS, this has been little studied and the mechanisms of IL-16 regulation in MS remain elusive . In this previous study, it was not apparent if IL-16 expression in MS differed from that of control adult CNS, or if the IL-16 expression was due to precursor or secreted IL-16. The cellular localization of IL-16 immunoreactivity suggested expression in microglia-like cells, although it did not correspond with the expression of several microglia activation markers. Some of the IL-16 immunoreactivity was found to be related to lymphocytes and astrocyte-like cells.
To asses the potential role of locally produced IL-16 in regulation of neuroinflammation and axonal damage, key markers of MS pathology, we specifically addressed the following questions: Does production of mature IL-16 occur within MS lesions? What are the phenotypes of IL-16-containing or -producing cells? Does intrathecal IL-16 accumulation correlate with specific CD4+ Th1 inflammation and damage of axonal cytoskeletons?
Here we present our findings on the expression and distinct regulation of pro- and secreted IL-16 in acute, subacute and chronic MS lesions, and in normal-appearing white matter, from brain and spinal cord of 39 patients with MS and 19 normal controls.
As anticipated, the highest levels of pro- and secreted IL-16 were found in acute lesions. The marked increases in secreted IL-16 over control levels in brain and spinal cord suggests a role for IL-16 in regulation of neuroinflammation in acute lesions. Locally produced IL-16 may further support production of inflammatory cytokines by infiltrating mononuclear cells, in a fashion similar to that seen in vitro . In this way, IL-16 may support inflammation in a complex network of pro- and anti- inflammatory pathways occurring in acute lesions [28, 29]. Surprisingly high levels of secreted IL-16 in chronic lesions, and lower but still above normal levels in normal appearing white matter, may be due to generalization of inflammatory and degenerative processes beyond focal demyelinating plaques and beyond macroscopically defined lesions. Such pathology is consistent with diffuse inflammation and with accumulating axonal injury during the progressive course of MS, as suggested by others . Alternatively, IL-16 found in chronic lesions and in normal white and grey matter adjacent to lesions may originate from neuronal IL-16 (NIL-16). Lymphoid IL-16 is transcribed from an intronic promoter lying within the NIL-16 gene. C-terminal regions of IL-16 and NIL-16 share 100% similarity. NIL-16 is produced by neurons primarily in cerebellum and in hippocampus. In vitro, at least, neuronal apoptosis leads to cleavage of NIL-16 by active caspase-3 and secretion of C-terminal bioactive IL-16 . Whether similar mechanisms operate in vivo during neuronal damage is not known. The lack of NIL-16 in normal control samples may reflect the anatomic distribution of NIL-16 expression. Alternatively, epitope for mAb 14.1 may be masked in NIL-16. Even if this antibody does detect intact NIL-16, following cleavage with active caspase-3, it would not detect the N-terminal domain of NIL-16. Our finding of IL-16 in MS tissue devoid of macroscopically significant mononuclear infiltration and/or microglial activation argues for a role for IL-16 in progression and spreading of focal inflammatory lesions. Our data showing substantially increased secreted IL-16 in CNS of mice with chronic EAE  support this proposed role for IL-16.
Slight differences were observed between brain and spinal cord in regulation of IL-16, active caspase-3, T-bet, Stat-2 (Tyr701) and NF(M+H)P, which may suggest the existence of distinct regulatory mechanisms of neuroinflammation in these two regions. The absence of appreciable IL-16 expression in control brain and spinal cord argues against an important physiological role for IL-16 in the maintenance of homeostasis within normal CNS.
In MS lesions, putative producers of IL-16 include CD4+, CD8+ T cells, B cells, dendritic cells, and subsets of activated microglia. Locally produced bioactive IL-16 may exert its paracrine effects by binding to CD4 co-receptor. A minimal peptide RRKS, between Arg106 and Ser109 of C-terminally cleaved bioactive IL-16, is essential for the binding and signaling through CD4 . This specific localization of intrathecal IL-16 bears relevance to its several potentially important roles: in local immunomodulation through regulation of CD4+T-T cell communication; in CD4+T cell helper function in the processes of T-B cell communication; and lastly in local antigen presentation by dendritic cells and microglia. A role for IL-16 in the cross talk between dendritic cells and T cells has also been proposed .
The Stat-1 protein plays a central role in the biological activity of both type I and type II interferons. Th1-derived IFNγ binds to its specific receptor on the surface of T cells and initiates signaling cascade, which includes Stat-1 phosphorylation and nuclear translocation. T-bet is regulated by IFNγ-, Stat-1-mediated signaling, and its level is markedly reduced in IFNγ- or Stat-1-mutant cells .
Inflammation-induced changes in axonal cytoskeletons, which include increased levels of phosphorylated medium and heavy chains of neurofilament, readily lead to axonal dysfunction [1, 35]. Our finding of IL-16 immunoreactivity in the vicinity of degenerate neurofilaments, which appears to be secreted from mononuclear cells (Fig. 6), suggests that this proinflammatory cytokine may participate in damage to axonal cytoskeletons.