In this study, we demonstrated that expression of endogenous SPI-3
is induced in specific cells of the ocular tissues in response to
inflammatory stimulation. The SPI-3 mRNA positive signal is found in
epithelial cells of the iris and ciliary body, and astrocytes in the
retina. Previous studies showed that SPI-3 is almost entirely silent in
normal (non-inflammatory) rats but is transiently activated in response
to inflammatory stimuli in the liver
16 18 and
pancreas.
31 The present study revealed a similar
inflammatory response of SPI-3 also exists in ocular tissues. This
transient expression is marked from 6 to 24 hours after LPS injection,
indicating that the expression occurred relatively early in the
inflammatory response. Thus, SPI-3 may inhibit proteolysis activity,
which occurs during the early phase of inflammation. Although the
function of this protease is still obscure, SPI-3 could be a marker for
the early phase of the inflammatory response. In the retina, the
inflammation elicits activation of some cell species, in particular,
microglia, macrophage, and Müller cells, and among these cells
little is known about the role of astrocytes under these conditions. In
this study, we demonstrate that among these reactive cell species, only
astrocytes can synthesize SPI-3, which may be one of their important
functions during inflammation. In addition, because SPI-3 has a
putative signal peptide sequence to be released and glycosylation is
evident,
18 this molecule is supposed to be released and
function in the extracellular space. In fact, all the SPI-3-mRNA
positive cells in the innermost layer of each tissue suggest that the
released SPI-3 can readily spread into the intraocular space. Because
there are no available antibodies to measure the released SPI-3, it is
impossible to confirm the release.
Previous studies speculated that because SPI-3 is expressed only under
inflammatory conditions, it might have protective effects against
inflammatory damage.
16 21 31 In fact, many serine protease
inhibitors have potent protective activities such as wound-healing
repairs. For instance, a recent paper reported that the secretory
leukocyte protease inhibitor (SLPI) is a pivotal endogenous factor
necessary for optimal wound healing.
32 SLPI is a
multi-potent serine protease inhibitor with anti-inflammatory,
anti-viral, anti-fungal and anti-bacterial properties. In addition,
SLPI antagonizes LPS-induced pro-inflammatory mediator synthesis by
monocytes and macrophages.
33 34 Although the functional
consequences of SPI-3 are yet unknown, it is likely that SPI-3 may have
similar anti-inflammatory properties under local inflammatory
conditions. Another intriguing aspect of inflammation concerns nerve
regeneration. Recently, Benowitz et al.
35 reported that
macrophage activation led to greatly increased regeneration of injured
optic nerves. They observed that the lens puncture somehow caused
massive infiltration of macrophage into the eye, which caused marked
activation of Müller cells. Under this inflammatory condition,
crushed optic nerve regeneration was significantly upregulated.
Although the molecular mechanism underlying how inflammation
accelerates nerve regeneration is unknown, an inflammation-activated
protease inhibitor such as SPI-3 might contribute to nerve regeneration
to some extent. Our unpublished data suggest that optic nerve injury
does not induce expression of SPI-3, whereas rat motor nerve
transection dramatically induced SPI-3 expression in the injured motor
neurons. Retinal ganglion cells with injured nerves cannot regenerate,
even though motor neurons with damaged nerves can survive and
regenerate, and the difference in SPI-3 expression may affect their
fate. Recently, the protein inhibitor 6 (PI-6; the human orthologue of
SPI-3) is proved as an inhibitor of the cathepsin G, and the cathepsin
G activates a proapoptotic protease, caspase-7.
36 37 Thus
PI-6 (or SPI-3) could be a potent inhibitor of caspase-7–mediated
apoptosis. In this respect SPI-3 may prevent a caspase-7–mediated
apoptosis, which is caused by damage such as nerve injury or
inflammatory stimulation. More detailed studies are need to determine
the functional significance of SPI-3.
In vitro studies have demonstrated that the expression of the SPI-3
gene is upregulated by IL-6.
19 20 22 23 This induction of
SPI-3 mRNA was observed approximately 1 hour after IL-6 stimulation in
rat-cultured primary hepatocytes, and the expression level peaked 24
hours after stimulation.
19 Coincidentally, it was reported
that expression of IL-6 mRNA significantly increases in the iris,
ciliary body, and retina in the early inflammatory phase of the rat EIU
model.
12 13 In addition, activation of the Janus kinase
(JAK)–STAT3 pathway in response to stress stimuli and ciliary
neurotrophic factor (CNTF) was also reported in retinal
astrocytes.
38 Because the STAT3 is rapidly activated by
phosphorylation at the tyrosine residue by JAKs after the IL-6
stimulation as well as CNTF, leukemia inhibitory factor, oncostatin M,
and IL-11,
39 40 41 we examined the STAT3 activation in the
retina after LPS injection using Western blot analysis and IHC. LPS
stimulation elicited STAT3 activation in retinal astrocytes. In
addition, the double-labeling study clearly demonstrated the
colocalization of pSTAT3 and SPI-3 in the astrocytes. An intriguing
aspect of this experiment is that the pSTAT3 appears slightly earlier
than SPI-3. Moreover, it was reported that the possible STAT3 binding
elements were included in the promoter region of the rat SPI-3 gene,
and the binding site was shown to be indispensable for SPI-3 induction
by IL-6.
20 These would strongly suggest that SPI-3
expression is induced after STAT3 activation in the retina of EIU
model. This in vivo result may support the in vitro evidence that SPI-3
expression occurs in a STAT3-mediated manner. Therefore, it might be
concluded that LPS elicits release of IL-6 and/or some
inflammation-associated cytokines from inflammatory cells such as a
macrophage or other cells and activates thereby the JAK-STAT pathways
in retinal astrocytes. This STAT3 activation might induce further SPI-3
expression to prevent a proteolytic insult caused by excessively
activated serine proteases.