Matrix metalloproteinases (
MMPs) that are secreted by activated T cells play a significant role in degradation of the extracellular matrix around the blood vessels and facilitate autoimmune
neuroinflammation; however, it remains unclear how
MMPs act in lesion formation and whether
MMP-targeted
therapies are effective in disease suppression. In the present study, we attempted to treat
experimental autoimmune encephalomyelitis (EAE) by administration of small interfering RNAs (siRNAs) for MMP-2, MMP-9, and
minocycline, all of which have
MMP-inhibiting functions.
Minocycline, but not siRNAs, significantly suppressed disease development. In situ zymography revealed that
gelatinase activities were almost completely suppressed in the spinal cords of
minocycline-treated animals, while significant
gelatinase activities were measured in the EAE lesions of control animals. However, MMP-2 and MMP-9 mRNAs and
proteins in the spinal cords of treated rats were unexpectedly upregulated. At the same time,
mRNA for tissue inhibitors of
MMPs (TIMP)-1 and -2 were also upregulated. The EnzChek
Gelatinase/
Collagenase assay using tissue containing native
MMPs and TIMPs demonstrated that
gelatinase activity levels in the spinal cords of treated rats were suppressed to the same level as those in normal spinal cord tissues. Finally, double immunofluorescent staining demonstrated that MMP-9 immunoreactivities of treated rats were almost the same as those of control rats and that MMP-9 and
TIMP-1 immunoreactivities were colocalized in the spinal cord. These findings suggest that
minocycline administration does not suppress
MMPs at
mRNA and
protein levels but that it suppresses
gelatinase activities by upregulating TIMPs. Thus,
MMP-targeted
therapies should be designed after the mechanisms of candidate drugs have been considered.