A large body of evidence suggests that ischemic preconditioning
renders various organs remarkably tolerant to ischemic
conditions.
1 5 Therefore, ischemic preconditioning has
attracted a great deal of attention, owing to its neuroprotective
property against focal and global cerebral ischemia.
24 25 In retina, Roth et al.
3 have recently reported the
neuroprotective effect of ischemic preconditioning against retinal
ischemia by means of histologic and functional analyses. Their findings
were supported by previous in vitro evidence that hypoxia increased
tolerance of retinal ganglion cells to anoxia.
26 In this
study, we investigated the effects of ischemic preconditioning on
leukocyte behavior during retinal ischemia–reperfusion injury, because
accumulating evidence has indicated that leukocytes play a central role
in postischemic neural damage.
8 9 10 Leukocytes that
accumulate in postischemic tissues have been suggested to cause injury
by blocking blood flow,
10 producing oxygen-free
radicals,
27 and releasing various types of inflammatory
cytokines.
28 In the present study, ischemic
preconditioning substantially inhibited leukocyte–endothelium
interactions in the postischemic retina. The inhibitory effects of
ischemic preconditioning on inflammatory leukocyte–endothelium
interactions in the postischemic retina would partially contribute to
the neuroprotective effect on the ischemic insult.
Recent experiments on leukocyte adhesion to the vascular endothelium
have shown that leukocyte recruitment to the area of inflammation takes
place through a multistep process mediated by distinct adhesion
molecules.
29 30 P-selectin–dependent leukocyte rolling is
a prerequisite to the establishment of intercellular adhesion molecule
(ICAM)-1–dependent adhesive interactions and subsequent leukocyte
emigration. We have reported that inhibition of P-selectin or ICAM-1
with the administration of monoclonal antibody substantially attenuates
leukocyte–endothelium interactions during retinal
ischemia–reperfusion injury.
20 Recently, Davis et
al.
31 have demonstrated the complete prevention of
postischemic P-selectin expression in rat jejunum by ischemic
preconditioning. In addition, ischemic preconditioning reportedly
reduces expression of ICAM-1 in cultured rat aortic endothelial cells
after anoxia-reoxygenation.
32 Retinal ischemic
preconditioning would suppress the expression of these adhesion
molecules during retinal ischemia–reperfusion injury, resulting in
attenuation of leukocyte rolling and subsequent leukocyte accumulation
in the postischemic retina.
Intense examinations of the mechanisms of ischemic preconditioning
indicate that adenosine plays a central role in ischemic tolerance
produced by preconditioning.
6 33 Although the precise
mechanism by which adenosine mediates preconditioning phenomenon is
uncertain, treatment with adenosine has been suggested to slow the rate
of metabolism and delay the accumulation of H
+ and Ca
+ during ischemia.
34 In the
present study, blocking of the adenosine A1 receptor by DPCPX resulted
in strong suppression of the inhibitory effects of ischemic
preconditioning on both leukocyte rolling and accumulation in
postischemic retina. Moreover, adenosine A1 receptor stimulation by
R-PIA produced partial but significant mimicking of the inhibition of
postischemic leukocyte behavior by ischemic preconditioning. Our
results show a central role of adenosine in reduced postischemic
leukocyte–endothelium cell interactions in preconditioned retinal
veins. In addition, adenosine has been known to act as an
anti-inflammatory molecule.
35 It has been suggested that
adenosine inhibits expression of adhesion molecules by activated
endothelial cells and, moreover, inhibits leukocyte adherence and
extravasation after ischemia-reperfusion.
36 It is feasible
that anti-inflammatory potential of adenosine would partially
contribute to reduced postischemic leukocyte–endothelium interactions
in preconditioned vessels.
In the present study, ischemic preconditioning strongly inhibited
leukocyte rolling and subsequent accumulation in the postischemic
retina. In the postischemic liver, a recent study using intravital
microscopy has reported that preconditioning attenuates
leukocyte–endothelium interactions in terminal hepatic
venules.
13 Moreover, Akimitsu et al.
11 have
reported the inhibitory effects of ischemic preconditioning on
postischemic leukocyte adhesion and emigration in skeletal muscle. They
also showed that adenosine may mediate the ability of ischemic
preconditioning to attenuate postischemic leukocyte–endothelium cell
interactions. Indeed, their results in the liver and the skeletal
muscle are compatible to our findings. However, Kubes et
al.
12 have shown that adenosine may play only a minor role
in reduced leukocyte–endothelium cell interactions in preconditioned
mesenteric venules after ischemia-reperfusion. They have also reported
that preconditioning had a minor effect on the flux of rolling
neutrophils in mesenteric venules after ischemia-reperfusion.
In the present study, ischemic preconditioning was induced in the
retina 24 hours before induction of prolonged ischemic insult. It has
been suggested that various organs may demonstrate different natures in
the preconditioning phenomenon. Ischemic preconditioning in the heart
induces protection in a biphasic pattern.
2 5 37 The early
preconditioning protective response is seen very early, lasting only
hours, and does not need protein synthesis; delayed preconditioning
phenomenon needs a day or a few days after ischemic preconditioning and
needs protein synthesis.
37 A report by Roth et
al.
3 has shown that preconditioning phenomenon in the
retina is not biphasic. Preconditioning before 24 or 72 hours before
ischemia completely prevents retinal damage, whereas a short time
interval between preconditioning and ischemia causes greater retinal
damage. Therefore, some difference in the nature of preconditioning
phenomenon between various organs may account for some discrepancy
between our findings and those in some previous reports.
In summary, the present study demonstrated the strong inhibitory
effects of ischemic preconditioning on leukocyte rolling and after
leukocyte accumulation in postischemic retina. In addition, adenosine
played an important role in inhibitory effects on
leukocyte–endothelium interactions through the A1 receptor. Retinal
ischemic preconditioning could partially exert neuroprotective effects
against prolonged ischemic insult by inhibition of
leukocyte–endothelium cell interactions through the adenosine A1
receptor.