Neuroscience Vol. 45, No. 1, pp. 221-225, 1991 Printed in Great Britain
0306-4522/91 $3.00 + 0.00
Perpmon Press plc 0 1991IBRO
PROSTAGLANDIN E, CHANGES IN THE RETINA AND OPTIC NERVE OF AN EYE WITH INJURED OPTIC NERVE A. BAR-ILAN,* N. NAvEH,*t
C. WEISSMAN,*M. BELRIN*and M. SCHWARTZ$
*Maurice and Gabriela Goldschleger Eye Research Institute, Tel Aviv University, Sackler Faculty of Medicine, Chaim Sheba Medical Center, Tel Hashomer, Israel SDepartment of Neurobiology, Weizmann Institute of Science, Rehovot, Israel Ahstract4hanges in arachidonic acid metabolism were studied in the optic nerve, the choriomtina, and in the vitreous following crush injury to the optic nerve of rats. Crush injury led to: (i) a 3.9-fold increase in optic nerve prostaglandin type Er in vitro production which peaked on day 5 and was followed by a gradual decline, but was still significantly higher than baseline levels by day 12; (ii) a two-fold increase in the chorioretina prostaglandin type E, in vitro production which peaked on day 1, and resumed baseline levels by day 3; (iii) a 3.5-fold increase in vitreous prostaglandin type E, levels on day 1 which remained at 1.5-2 times higher than baseline levels for the rest of the study period (12 days). The findings indicate that the pattern of changes in prostaglandin type E, production by the optic nerve (consisting mostly of white matter) is different from that described for injured brain tissues. The prolonged accumulation of vitreal prostaglandin type E, in eyes with damaged optic nerve may lead to undesirable effects on the retina beyond those directly manifested in the retina by altered axonal flow in the injured optic nerve.
Injury to the optic nerve, both direct and indirect, as in the adjacent vitreous. Pharmacological manipuusually results in immediate total loss of vision. Both lation of such biochemical changes may serve as a surgical optic nerve decompression and high dose basis for medical therapy in cases of optic nerve corticosteroids have been used therapeutically, with injury. varying degrees of success. The efficacy of these EXPERIMENTAL PRDCEDURRS treatments is currently unknown, as no prospective Optic nerve crush injury was performed in male albino controlled studies have been carried out comparing rats (Charles River, T.A.U. Animal Resources Unit), the two treatments with no treatment.22*27*32s3v 0.24.3 kg, anesthetized with an intraneritoneal iniection of Prostaglandins and other eicosanoids are syn40 mg/kgketamine HCl (Vetalar, Parke-Davis) and 8 mg/kg thesized through the cyclooxygenase pathway from xylazine-HCl (Rompum, Bayer, Leverkusen). After incision arachidonic acid, which is released from cell mem- of the conjunctiva and displacement of the ocular muscles brane phospholipids whenever the cells are exposed the optic nerve was exposed and crushed for 3Os, 2 mm distal to the eyeball, using a reverse action capsule forceps to various stimuli, such as hypoxia or mechanical modified to enable adjustment of compression power. The stimulation.38 Prostaglandins are also known for their pressure applied was predetermined to cause an immediate mediatory role in the reaction to injury.v~24An im- disappearance of the compound action potential followed by its recovery 30min later to 200-5OOuV.’ The animals proved neurological recovery was reported in animals suffering from spinal trauma or brain ischemia fol- were divided into two groups: (i) crush’injury group; (ii) sham-operated group (animals that underwent injury as in lowing a combined drug treatment that included the group i, but without crushing the nerves). The rats were cyclooxygenase inhibitor indomethacin.i4 There is killed at zero time (immediately after surgery, nine only sparse information on the involvement of pros- rats/group), and one, three, five, and 12 days later (seven taglandins in retinal pathophysiology: the retina ex- rats/group), by an overdose of pentobarbitone. The optic hibits cyclooxygenase activity,36 is directly affected by nerve was exposed, and detached from the eyeball following skull opening. This allowed dissection of the nerve at the excessive prostaglandins as shown by electro-retinchiasm, so that the whole length of the nerve could be E2 obtained. Much care was taken to minimize trauma to the ography,43 whereas an increased prostaglandin (PGEJ production by the chorioretina was observed optic nerve during tissue detection, as this can greatly al&t following retinal injury induced by laser ir- eicosanoid production. ‘iJ6 The enucleated eyes were frozen in liquid nitrogen and dissected later. The anterior segment radiation.33*u of the eye was removed, while the vitreous and the whole We therefore decided to study the changes in retina attached to the choroid were each carefully separated arachidonic acid metabolism in injured optic nerve, from the sclera, and placed in separate vials. Weighed optic and the corresponding changes in the retina as well nerve sections, vitreous, and chorioretina were each incubated separately in Krebs-Ringer-bicarbonate at 37°C for 30 and 10 min, respectively. In a few cases when retina tissue weight was too low (c 10 mg), tissues from two eyes were pooled and assayed together. Incubation media PGE, levels
tTo whom correspondence should be addressed. Abbreuiations: PGE,, prostaglandin type E,. 221
A. BAK-IL.AN
222 optic
OJ :
I
:
01
I
:
:
time
Fig. 1.Changes in PGEz following crush injury. (control); e-0, optic different from control
nerve
:
:
:
:
:
:
5
3
er ul
:
:
J
12
04 :
I
o
1
:
:
:
I
time
(days)
in oitro production by optic nerves O-0, Sham-operated animals nerve crush injury. Significantly at: * P < 0.05: *** P < 0.005.
were determined using a radioimmunoassay
with specific antibody to PGE, (Miles Yeda, Rehovot), with a 3% cross-reactivity with PGE,. The cross-reactivity with other prostaglandins was less than 3%. PGE, levels in the incu-
bation media containing either the optic nerve or the chorioretina represent the amounts produced in vitro by the appropriate tissue. PGEz detected in the vitreous sample originated from excessive in vivochorioretinal production as PGE,s are not stored intracellularly, but are immediately released into the adjacent extracellular space, mostly the vitreous cavity. Data (mean f S.E.) of PGE, levels are expressed as pg/mg
wet tissue weight. Analysis of variance (ANOVA) and Student’s r-test were used to calculate the significance of differences between the various groups.@
:
/
5
!
I
I
)
:
t-
12 (doys)
Fig. 3. Changes in PGE, in the vitreous following optic nerve crush injury. Symbols as in Fig. 1. Significantly different from control at: ** P -c0.01: ***P < 0.005.
levels were still significantly higher than baseline levels (P < 0.05, Fig. 1). The average chorioretinal PGEz in vitro production in sham-operated (control) eyes was 13.5 * 1.2 pg/mg. Crush injury to the optic nerve led to a transient two-fold increase in the chorioretina PGE, production, which peaked on day 1 (200% of control, P < O.Ol), and resumed baseline levels by day 3 (Fig. 2). Vitreous PGEz concentration levels, which were 2.1 + 0.4 pg/mg in sham-operated animals, showed a 3.5-fold increase on day 1 (P < 0.01) followed by a decline, but remained elevated above baseline levels for the rest of the study period (up to day 12, Fig. 3).
RESULTS
The average optic nerve PGE, in vitro production in sham-operated (control) animals was 100 + 17 pg/mg (n = 9). Crush injury resulted in a gradual increase in the optic nerve in vitro PGE, production
DISCUSSION Our finding of a two- to four-fold increase PGE, in vitro production by the injured optic nerve, as well as
levels, which peaked on day 5 (3.9 times higher than control levels, P < 0.01). The ensuing decline in PGEz levels was at a slower rate (3O%/day vs 80%/day during the elevation phase), and by day 12 PGEz chorio-retina 30
o-0 .-.
25
5J
: 0
i 1
:
: 3
I
: 5
time
:
:
:
:
control crush injury
j
:
: + 12
1
(days)
Fig. 2. Changes in PGEl in vitro production by the chorioretina following optic nerve crush injury. Symbols as in Fig. 1. Significantly different from control at: * P i 0.05; *** P < 0.005.
the levels found in our normal optic nerve, are consistent with previous reports on traumatized CNS,9,‘5,21and normal brain tissues.23 On the other hand, the temporal pattern of the enhanced PGE, in vitro production in the crushed optic nerve is strikingly different from that reported in earlier studies of brain and spinal cord injury. The increased levels of cyclooxygenase products reported for traumatized brain tissues peaked soon after injury, resuming baseline levels within a few hours after perturbation,912’@’with a second peak observed in some brain areas 24-72 h following brain ischemia.“* On the other hand, in the current study the enhanced production was maintained over much longer periods (up to 12 days). This variance in the temporal changes in PGE, production might be due to differences in PGE, synthesis capacity and/or the mechanisms of its elimination between the optic nerve (consisting mostly of white matter), and various brain areas and spinal cord (containing both white and gray matter). Evidence for differences between PGE, production by the brain’s gray and white matter in response to various stimuli, such as hemorrhagic hypotension,
Prostaglandin E, changes in the retina and optic nerve of the eye with injured optic nerve
acute hypoxemia, and severe asphyxia has already been observed in the beagle pup model of brain injury.28+29,31 Moreover, the brain’s gray and white matter differ also in their blood flow responsiveness to acute hypoxemia.30 Various changes that occur in injured nerves, including selective nerve fiber degeneration, monolipid cytes/macrophages infiltration, altered metabolism, etc. were reported to appear two to three days after injury and span over a period of a few weeks.8*4’,45,47 The temporal pattern of optic nerve PGE, production changes coincides with these changes but the significance of these changes in the degeneration-regeneration-related processes are yet unknown. The development of post-trauma brain tissue edema was reported to be associated with changes in PGE,.4,19 Our observation of an enhanced PGE, in vitro production by the injured nerve may be partially related to the marked edema evident in optic nerve following crush injury.42 The finding of an enhanced chorioretinal PGE, production in eyes subjected to optic nerve injury indicates that trauma to optic nerve axons may affect the cell bodies located in the neuroretina, partially through an altered retrograde axonal transport.” Mechanisms leading to the chorioretinal PGE, production changes by traumatized optic nerve are as yet unknown. Our method does not allow us to determine whether the retina, choroid, or both are responsible for the enhanced production. However, retinal injury induced by laser irradiation which involves both the retina and choroidal layer were associated with an increased chorioretinal PGE, in vitro production.” It is important to note that the chorioretinal changes in PGE, production were not closely associated with those observed in the damaged optic nerve: the maximal chorioretinal response occurred on day 1 (vs day 5 in the optic nerve), and it resumed baseline levels on day 3 (vs the extended elevation for at least 12 days in the optic nerve). Chorioretinal PGE, changes in eyes with injured nerves were of shorter duration (three days) than those observed when injury was directly inflicted to both layers by laser irradiation (persisting for 14 days).33 The mechanisms underlying these differences remain to be unveiled, and are probably related to the site and nature of the damage, and to the composition of the tissues involved (mostly nerve fibers in the optic nerve vs a variety of nerve cell bodies plus fibers in the retina. Vitreal PGE2 content is determined by the balance between the amounts of PGE2 produced by the chorioretina, which are immediately released into the
223
vitreous, and the rate of their removal from this extracellular space by various transport mechanisms located in the ciliary body and the blood-retinal barrier.536 Following optic nerve injury the initial increase in vitreal PGE, levels corresponded well with PGE, chorioretinal peak levels, whereas later (beyond day 5) the extended elevation in vitreal PGE2 levels was maintained despite the reduced chorioretinal PGE, production at that period. Elevated vitreal PGE, concurrent with diminishing chorioretinal in vitro PGE, formation have already been observed previously in rabbits’ eyes exposed to Neodynium: Yag laser,33~34a discrepancy attributed to a derangement in vitreal PGE, removal mechanism. The prolonged elevation in vitreal PGE, levels in our injured optic nerve might be due to an altered removal mechanism or additionally represent a continuous entry of PGE, from areas other than the chorioretina (i.e. the optic nerve head). Our current findings of long-term elevation of PGE, in the optic nerve and retina indicate a longterm inflammatory reaction in these tissues following optic nerve crush injury. Although prostaglandins are well known as mediators of inflammation, and are synthesized in injured irritated tissueq3’@ their specific role in the complex inflammatory response is not fully known. Their involvement in a variety of biological processes related to tissue degradation and reconstruction like cell proliferation, protein metabolism etc. 217~12,20,25,26,48 some of which have been observed in ‘the injured optic nerve and retina,4’.42 suggests a multifaceted role. CONCLUSIONS
In summary: (i) our demonstration that optic nerve crush injury in rats’ eyes is associated with a pronounced and lengthy increase in PGE2 in vitro production by the traumatized nerve indicates a long-term reaction; (ii) the white matter temporal PGE, response to various non-chemical stimuli may be different from that observed in brain tissue; (iii) the prolonged accumulation of PGE2 in the vitreous of eyes with injured optic nerve may in itself have undesirable effects on the retina35s43beyond those attributable to the transient optic nerve injury. Further studies correlating prostaglandin (and other eicosanoid) levels, specific inflammatory processes, and concomitant loss of visual function would help in defining the therapeutic potential and clinical applicability of the current findings. Acknowledgement-This work was partially supported by the Gotthelf Foundation of the Tel Aviv University.
REFERENCES
1. Abe K., Yoshidomi M. and Kogure K. (1989) Arachidonic acid metabolism in ischemic neuronal damage. Ann. N. Y. Acad. Sci. 569, 259-268. 2. Akatsu T., Takahashi N., Debari K., Morita I., Murota S., Nagata N., Takatani 0. and Suda T. (1989) Prostaglandins NSC45,1--n
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 21. 28.
29. 30. 31 32 33
34. 35. 36. 37. 38.
promote osteoclast-like cell formation by a mechanism mvolving cyclic adinosine 3,5-monophosphate in mouse hon,~ marrow cell cultures. J. Bone Min. Res. 4, 29935. Assia E., Rosner M., Belkin M.. Solomon A. and Schwartz M. (1989) Temporal parameters of low energy iasei irradiation for optimal delay of post-traumatic degeneration of rat optic nerve. Bruin Res. 476, 205.212. Bhakoo K. K., Crockard H. A.. Lascelles P. C. and Avery S. F. (1984) Prostaglandin synthesis and oedema formation during reperfusion following experimental brain ischaemia in the gerbil. Stroke 15, 891.895. Bhattacherjee P. (1974) Autoradiographic localization of intravitreally or intracamerally injected (‘H) prosraglandin. Expl Eye Res. 18, 181-188. Bito L. Z. and Salvador E. V. (1972) Intraocular fluid dynamics: III. The site and mechanisms of prostaglandins transfer across the blood intraocular fluid barriers. Exp/ Eye Res. 14, 233-241. Bitterman P. B., Wewers M. D., Rennard S. I., Adelberg S. and Crystal R. G. (1986) Modulation of alveolar macrophages-derived fibroblast proliferation by alternative macrophage mediators. J. clin. Inuest. 77, 700~708. Burke W., Cotte L. J., Garvey J., Kumarasinghe R. and Kyniacau C. (1986) Selective degeneration of optic nerve fibers in the cat produced by pressure block. J. Physiol. 376, 461476. Dempsey R. J., Roy M. W., Meyer K., Cowen D. E. and Tai H. H. (1986) Development of cyclooxygenase and lipoxygenase metabolites of arachidonic acid after transient cerebral ischemia. J. Neurosurg. 64, 118-124. Feist R. M., Kline L. B., Morris R. E., Witherspoon C. D. and Michelson M. A. (1987) Recovery of vision after presumed direct optic nerve injury. Ophthalamology 94, I567- 1569. Feldman S. T., Rodriguez P., Cabrera A. C. and Mitchell M. (1990) The effect of temperature and trephination on eicosanoid release by donor corneas. Inoest. Ophlhalm. vis. Sci. 31, 486. Fitzpatrick L. R., Gaginella T. B., Haddox M. K. and Johnson L. R. (1988) Prostaglandin-mediated trophic effects on the rat duodenum: the role of polyamines. Proc. Sot. exp. Biol. Med. 189, 201-205. Glazer J. S. (1988) Topical diagnosis: prechiasmal visual pathways. In Clinical Ophthalmology (eds Duane T. D. and Jaeger E. A.), Vol. 2, pp. 65-66. J. B. Lippincott, PA. Hallenbeck J. M., Jacobs T. P. and Faden A. 1. (1983) Combined PGE,, indomethacin and heparin improves neurological recovery after spinal &hernia in cats. J. Neurosurg. 58, 7499754. Hanamura T., Shigeno T.. Asano T., Mima T. and Takakura K. (1989) Prostaglandin profiles in relation to local circulatory changes following focal cerebral ischemia in cats. Stroke 20, 803-808. Hotter G., Roseho-Catafu J.. Bulbena 0.. Gomez G., Colomer J., Saenz F. P. A., Cruz L. F. and Gelpi E. (1990) Prostaglandin E, and thromboxane Bz levels in rats subjected to pancreas transplantation. Prostaglandins 39, 5360. Hunt A. H. and Minckler D. S. (1983) Optic nerve axonal transport: basic aspects. In Biomedical Foundations of Ophthalmology (eds Duane T. D. and Jaeger E. A.), Vol. 1, Chap. 26. Harper and Row, PA. Hgu C. Y., Liu T. H., Yu J., Hogan E. L., Chao J., Sun G., Tai H. H., Beckman J. S. and Freman B. A. (1989) Arachidonic acid and its metabolites in cerebral ischemia. Ann. N. Y. Acud. Sci. 559, 282-295. Ianotti F., Crockard A., Ladds G. and Symon L. (1981) Are prostaglandins involved in experimental ischemic edema in gerbils. Stroke 12, 301-306. Johnson G. S. and Pastan I. (1971) Changes in growth and morphology of fibroblasts by prostaglandins. J. nam. Cancer Res. 47, 1357-1364. Jonsson H. T. and Daniel1 H. B. (1976) Altered levels of PGF in cat spinal cord tissue following traumatic injury. Prostaglandins 11, 5 l-6 1. Joseph M. P.. Lessel S., Rizzo J. and Momose J. (1990) Extracranial optic nerve decompression for traumatic optic neuropathy. Arch. Ophrhal. 108, 1091~-1093. Katz B., Sofonio M., Lyden P. D. and Mitchell M. D. (1988) Prostaglandin concentrations in cerebrospinal fluid of rabbits under normal and ischemic conditions. Stroke 19, 349-351. Leslie J. B. and Watkin W. D. (1985) Eicosanoids in the central nervous system. J. Neurosurg. 63, 6599668. Libby P., Warner S. J. C. and Friedman G. B. (1988) Interleukin I: a mitogen for human vascular smooth muscle cells that induces the release of growth-inhibitory prostanoids. J. clin. Invest. 81, 487498. Logn J. L., Lee S. M., Benson B. and Michael U. F. (1986) Inhibition of compensatory renal growth of indomethacin. Prostaglandins 31, 2533261. Marouf L. and Azar D. T. (1985) Reversible visual loss following optic nerve injury. Ann. Ophfhal. 17, 5822584. Ment L. R., Stewart W. B., Duncan C. C., Cole J. S. and Pitt B. R. (1985) Beagle puppy model of perinatal cerebral infarction. Acute changes in regional cerebral prostaglandins during hemorrhagic hypotension. J. Neurosurg. 63, 899-904. Ment L. R., Stewart W. B., Duncan C. C. and Pitt B. R. (1986) Beagle puppy model of perinatal cerebral insults. Cerebral blood flow changes and intraventricular hemorrhage evoked by hypoxemia. J. Neurosurg. 65, 847.-850. Ment L. R., Stewart W. B., Duncan C. C., Pitt B. R. and Cole J. S. (1986) Beagle puppy model of perinatal cerebral infarction. Regional cerebral prostagiandins changes during acute hypoxemia. J. Neurosurg. 65, 851-855. Ment L. R., Stewart W. B., Duncan C. C., Pitt B. R. and Cole J. S. (1987) Beagle pup model of brain injury: regional cerebral blood flow and cerebral prostaglandins. J. Neurosurg. 67, 2788283. Miller N. R. (1990) The management of traumatic optic neuropathy. Arch. Ophthal. 108, 10861087. Naveh N., Rosner M., Zborowsky-Gutman L., Rosen N. and Weissman C. (1990) A comparison of Argon and Neodynium: YAG laser iridotomy effect on prostaglandins E, and blood aqueous barrier disruption. Ophrhal. Res. 22, 253-259. Naveh N. and Weissman C. (1990) The correlation between excessive vitreal protein levels and prostaglandins PGE, levels on the blood retinal barrier. Prostaglundins 39, 147-156. Peyman G. A., Bennet T. D. and Velchek i (1975) The effects of intravitreal prostaglandin on retinal vasculature. Ann. Ophthal. 7, 7299134. PreudHomme Y., Demolle D. and Boeynaems J. M. (1985) Metabolism of arachidonic acid in rabbit iris and retina. Invest. Ophthal. vis. Sci. 26, 13361342. Salmon J. A. and Higgs G. A. (1987) Prostaglandins and leukotrienes as inflammatory mediators. Br. med. Bull. 43, 2855296. Samuelson B. (1987) An elucidation of the arachidonic cascade. Drugs 33, Suppl. 1, 2-9.
Prostaglandin
E, changes in the retina and optic nerve of the eye with injured optic nerve
225
39. Seiff S. R. (1990) High dose corticosteroids for treatment of vision loss due to indirect injury to the optic nerve. Ophthal. Surg. 21, 389-395. 40. Shohami E., Shapira Y., Sidi A. and Cotev S. (1987) Head injury induces increased prostaglandin synthesis in rat brain. J. cerebr. Blood Flow Metab. 7, 5863. 41. Shwartx M. (1987) Molecular and cellular aspects of nerve regeneration. CRC Biochem. 22, 89-110. 42. Shyne-Athwal S., Chakraborty G., Gage E. and Ingoglia N. A. (1988) Comparison of posttranslational protein modification by amino acid addition after crush injury to sciatic and optic nerves of rats. Expi Neurol. 99, 281-295. 43. Siminoff R. and Bito L. Z. (1982) The effects of prostaglandins and arachidonic acid on the electroretinogram: evidence for functional cyclooxygenase activity in the retina. Curr. Eye Res. 1, 635-642. 44. Sokal R. R. and RohIf F. J. (1969) Biometry, the principles and practice of statistics in biological research. W. H.
Freeman, San Francisco. 45. Stoll G., Trapp B. D. and Griffin J. W. (1989) Macrophage function during Wallerian degeneration of rat optic nerve: clearance of degenerating myelin and Ia expression. J. Neurosci. 9, 2327-2335. 46. Weissmann G. (1987) Pathogenesis of inflammation. Effects of the pharmacological manipulation of arachidonic acid metabolism on the cytological response to inflammatory stimuli. Drugs 33, 28-37. 47. White F. V., Toews A. D., Goodman J. F., Novicki D. L., Bouldin T. W. and Morel1 P. (1989) Lipid metabolism during early stages of Wallerian degeneration in the rat sciatic nerve. J. Neurochem. 52, 108>1092. 48. Wiley M. H., Feingold K. R., Grunfeld C., Quensey-Hunbeus V. and Wu J. M. (1983) Evidence for CAMP-independent inhibition of S-phase DNA synthesis by prostaglandins. J. biol. Chem. 258, 491496. (Accepted
7 May
1991)
0306-4522/91 $3.00 + 0.00
Perpmon Press plc 0 1991IBRO
PROSTAGLANDIN E, CHANGES IN THE RETINA AND OPTIC NERVE OF AN EYE WITH INJURED OPTIC NERVE A. BAR-ILAN,* N. NAvEH,*t
C. WEISSMAN,*M. BELRIN*and M. SCHWARTZ$
*Maurice and Gabriela Goldschleger Eye Research Institute, Tel Aviv University, Sackler Faculty of Medicine, Chaim Sheba Medical Center, Tel Hashomer, Israel SDepartment of Neurobiology, Weizmann Institute of Science, Rehovot, Israel Ahstract4hanges in arachidonic acid metabolism were studied in the optic nerve, the choriomtina, and in the vitreous following crush injury to the optic nerve of rats. Crush injury led to: (i) a 3.9-fold increase in optic nerve prostaglandin type Er in vitro production which peaked on day 5 and was followed by a gradual decline, but was still significantly higher than baseline levels by day 12; (ii) a two-fold increase in the chorioretina prostaglandin type E, in vitro production which peaked on day 1, and resumed baseline levels by day 3; (iii) a 3.5-fold increase in vitreous prostaglandin type E, levels on day 1 which remained at 1.5-2 times higher than baseline levels for the rest of the study period (12 days). The findings indicate that the pattern of changes in prostaglandin type E, production by the optic nerve (consisting mostly of white matter) is different from that described for injured brain tissues. The prolonged accumulation of vitreal prostaglandin type E, in eyes with damaged optic nerve may lead to undesirable effects on the retina beyond those directly manifested in the retina by altered axonal flow in the injured optic nerve.
Injury to the optic nerve, both direct and indirect, as in the adjacent vitreous. Pharmacological manipuusually results in immediate total loss of vision. Both lation of such biochemical changes may serve as a surgical optic nerve decompression and high dose basis for medical therapy in cases of optic nerve corticosteroids have been used therapeutically, with injury. varying degrees of success. The efficacy of these EXPERIMENTAL PRDCEDURRS treatments is currently unknown, as no prospective Optic nerve crush injury was performed in male albino controlled studies have been carried out comparing rats (Charles River, T.A.U. Animal Resources Unit), the two treatments with no treatment.22*27*32s3v 0.24.3 kg, anesthetized with an intraneritoneal iniection of Prostaglandins and other eicosanoids are syn40 mg/kgketamine HCl (Vetalar, Parke-Davis) and 8 mg/kg thesized through the cyclooxygenase pathway from xylazine-HCl (Rompum, Bayer, Leverkusen). After incision arachidonic acid, which is released from cell mem- of the conjunctiva and displacement of the ocular muscles brane phospholipids whenever the cells are exposed the optic nerve was exposed and crushed for 3Os, 2 mm distal to the eyeball, using a reverse action capsule forceps to various stimuli, such as hypoxia or mechanical modified to enable adjustment of compression power. The stimulation.38 Prostaglandins are also known for their pressure applied was predetermined to cause an immediate mediatory role in the reaction to injury.v~24An im- disappearance of the compound action potential followed by its recovery 30min later to 200-5OOuV.’ The animals proved neurological recovery was reported in animals suffering from spinal trauma or brain ischemia fol- were divided into two groups: (i) crush’injury group; (ii) sham-operated group (animals that underwent injury as in lowing a combined drug treatment that included the group i, but without crushing the nerves). The rats were cyclooxygenase inhibitor indomethacin.i4 There is killed at zero time (immediately after surgery, nine only sparse information on the involvement of pros- rats/group), and one, three, five, and 12 days later (seven taglandins in retinal pathophysiology: the retina ex- rats/group), by an overdose of pentobarbitone. The optic hibits cyclooxygenase activity,36 is directly affected by nerve was exposed, and detached from the eyeball following skull opening. This allowed dissection of the nerve at the excessive prostaglandins as shown by electro-retinchiasm, so that the whole length of the nerve could be E2 obtained. Much care was taken to minimize trauma to the ography,43 whereas an increased prostaglandin (PGEJ production by the chorioretina was observed optic nerve during tissue detection, as this can greatly al&t following retinal injury induced by laser ir- eicosanoid production. ‘iJ6 The enucleated eyes were frozen in liquid nitrogen and dissected later. The anterior segment radiation.33*u of the eye was removed, while the vitreous and the whole We therefore decided to study the changes in retina attached to the choroid were each carefully separated arachidonic acid metabolism in injured optic nerve, from the sclera, and placed in separate vials. Weighed optic and the corresponding changes in the retina as well nerve sections, vitreous, and chorioretina were each incubated separately in Krebs-Ringer-bicarbonate at 37°C for 30 and 10 min, respectively. In a few cases when retina tissue weight was too low (c 10 mg), tissues from two eyes were pooled and assayed together. Incubation media PGE, levels
tTo whom correspondence should be addressed. Abbreuiations: PGE,, prostaglandin type E,. 221
A. BAK-IL.AN
222 optic
OJ :
I
:
01
I
:
:
time
Fig. 1.Changes in PGEz following crush injury. (control); e-0, optic different from control
nerve
:
:
:
:
:
:
5
3
er ul
:
:
J
12
04 :
I
o
1
:
:
:
I
time
(days)
in oitro production by optic nerves O-0, Sham-operated animals nerve crush injury. Significantly at: * P < 0.05: *** P < 0.005.
were determined using a radioimmunoassay
with specific antibody to PGE, (Miles Yeda, Rehovot), with a 3% cross-reactivity with PGE,. The cross-reactivity with other prostaglandins was less than 3%. PGE, levels in the incu-
bation media containing either the optic nerve or the chorioretina represent the amounts produced in vitro by the appropriate tissue. PGEz detected in the vitreous sample originated from excessive in vivochorioretinal production as PGE,s are not stored intracellularly, but are immediately released into the adjacent extracellular space, mostly the vitreous cavity. Data (mean f S.E.) of PGE, levels are expressed as pg/mg
wet tissue weight. Analysis of variance (ANOVA) and Student’s r-test were used to calculate the significance of differences between the various groups.@
:
/
5
!
I
I
)
:
t-
12 (doys)
Fig. 3. Changes in PGE, in the vitreous following optic nerve crush injury. Symbols as in Fig. 1. Significantly different from control at: ** P -c0.01: ***P < 0.005.
levels were still significantly higher than baseline levels (P < 0.05, Fig. 1). The average chorioretinal PGEz in vitro production in sham-operated (control) eyes was 13.5 * 1.2 pg/mg. Crush injury to the optic nerve led to a transient two-fold increase in the chorioretina PGE, production, which peaked on day 1 (200% of control, P < O.Ol), and resumed baseline levels by day 3 (Fig. 2). Vitreous PGEz concentration levels, which were 2.1 + 0.4 pg/mg in sham-operated animals, showed a 3.5-fold increase on day 1 (P < 0.01) followed by a decline, but remained elevated above baseline levels for the rest of the study period (up to day 12, Fig. 3).
RESULTS
The average optic nerve PGE, in vitro production in sham-operated (control) animals was 100 + 17 pg/mg (n = 9). Crush injury resulted in a gradual increase in the optic nerve in vitro PGE, production
DISCUSSION Our finding of a two- to four-fold increase PGE, in vitro production by the injured optic nerve, as well as
levels, which peaked on day 5 (3.9 times higher than control levels, P < 0.01). The ensuing decline in PGEz levels was at a slower rate (3O%/day vs 80%/day during the elevation phase), and by day 12 PGEz chorio-retina 30
o-0 .-.
25
5J
: 0
i 1
:
: 3
I
: 5
time
:
:
:
:
control crush injury
j
:
: + 12
1
(days)
Fig. 2. Changes in PGEl in vitro production by the chorioretina following optic nerve crush injury. Symbols as in Fig. 1. Significantly different from control at: * P i 0.05; *** P < 0.005.
the levels found in our normal optic nerve, are consistent with previous reports on traumatized CNS,9,‘5,21and normal brain tissues.23 On the other hand, the temporal pattern of the enhanced PGE, in vitro production in the crushed optic nerve is strikingly different from that reported in earlier studies of brain and spinal cord injury. The increased levels of cyclooxygenase products reported for traumatized brain tissues peaked soon after injury, resuming baseline levels within a few hours after perturbation,912’@’with a second peak observed in some brain areas 24-72 h following brain ischemia.“* On the other hand, in the current study the enhanced production was maintained over much longer periods (up to 12 days). This variance in the temporal changes in PGE, production might be due to differences in PGE, synthesis capacity and/or the mechanisms of its elimination between the optic nerve (consisting mostly of white matter), and various brain areas and spinal cord (containing both white and gray matter). Evidence for differences between PGE, production by the brain’s gray and white matter in response to various stimuli, such as hemorrhagic hypotension,
Prostaglandin E, changes in the retina and optic nerve of the eye with injured optic nerve
acute hypoxemia, and severe asphyxia has already been observed in the beagle pup model of brain injury.28+29,31 Moreover, the brain’s gray and white matter differ also in their blood flow responsiveness to acute hypoxemia.30 Various changes that occur in injured nerves, including selective nerve fiber degeneration, monolipid cytes/macrophages infiltration, altered metabolism, etc. were reported to appear two to three days after injury and span over a period of a few weeks.8*4’,45,47 The temporal pattern of optic nerve PGE, production changes coincides with these changes but the significance of these changes in the degeneration-regeneration-related processes are yet unknown. The development of post-trauma brain tissue edema was reported to be associated with changes in PGE,.4,19 Our observation of an enhanced PGE, in vitro production by the injured nerve may be partially related to the marked edema evident in optic nerve following crush injury.42 The finding of an enhanced chorioretinal PGE, production in eyes subjected to optic nerve injury indicates that trauma to optic nerve axons may affect the cell bodies located in the neuroretina, partially through an altered retrograde axonal transport.” Mechanisms leading to the chorioretinal PGE, production changes by traumatized optic nerve are as yet unknown. Our method does not allow us to determine whether the retina, choroid, or both are responsible for the enhanced production. However, retinal injury induced by laser irradiation which involves both the retina and choroidal layer were associated with an increased chorioretinal PGE, in vitro production.” It is important to note that the chorioretinal changes in PGE, production were not closely associated with those observed in the damaged optic nerve: the maximal chorioretinal response occurred on day 1 (vs day 5 in the optic nerve), and it resumed baseline levels on day 3 (vs the extended elevation for at least 12 days in the optic nerve). Chorioretinal PGE, changes in eyes with injured nerves were of shorter duration (three days) than those observed when injury was directly inflicted to both layers by laser irradiation (persisting for 14 days).33 The mechanisms underlying these differences remain to be unveiled, and are probably related to the site and nature of the damage, and to the composition of the tissues involved (mostly nerve fibers in the optic nerve vs a variety of nerve cell bodies plus fibers in the retina. Vitreal PGE2 content is determined by the balance between the amounts of PGE2 produced by the chorioretina, which are immediately released into the
223
vitreous, and the rate of their removal from this extracellular space by various transport mechanisms located in the ciliary body and the blood-retinal barrier.536 Following optic nerve injury the initial increase in vitreal PGE, levels corresponded well with PGE, chorioretinal peak levels, whereas later (beyond day 5) the extended elevation in vitreal PGE2 levels was maintained despite the reduced chorioretinal PGE, production at that period. Elevated vitreal PGE, concurrent with diminishing chorioretinal in vitro PGE, formation have already been observed previously in rabbits’ eyes exposed to Neodynium: Yag laser,33~34a discrepancy attributed to a derangement in vitreal PGE, removal mechanism. The prolonged elevation in vitreal PGE, levels in our injured optic nerve might be due to an altered removal mechanism or additionally represent a continuous entry of PGE, from areas other than the chorioretina (i.e. the optic nerve head). Our current findings of long-term elevation of PGE, in the optic nerve and retina indicate a longterm inflammatory reaction in these tissues following optic nerve crush injury. Although prostaglandins are well known as mediators of inflammation, and are synthesized in injured irritated tissueq3’@ their specific role in the complex inflammatory response is not fully known. Their involvement in a variety of biological processes related to tissue degradation and reconstruction like cell proliferation, protein metabolism etc. 217~12,20,25,26,48 some of which have been observed in ‘the injured optic nerve and retina,4’.42 suggests a multifaceted role. CONCLUSIONS
In summary: (i) our demonstration that optic nerve crush injury in rats’ eyes is associated with a pronounced and lengthy increase in PGE2 in vitro production by the traumatized nerve indicates a long-term reaction; (ii) the white matter temporal PGE, response to various non-chemical stimuli may be different from that observed in brain tissue; (iii) the prolonged accumulation of PGE2 in the vitreous of eyes with injured optic nerve may in itself have undesirable effects on the retina35s43beyond those attributable to the transient optic nerve injury. Further studies correlating prostaglandin (and other eicosanoid) levels, specific inflammatory processes, and concomitant loss of visual function would help in defining the therapeutic potential and clinical applicability of the current findings. Acknowledgement-This work was partially supported by the Gotthelf Foundation of the Tel Aviv University.
REFERENCES
1. Abe K., Yoshidomi M. and Kogure K. (1989) Arachidonic acid metabolism in ischemic neuronal damage. Ann. N. Y. Acad. Sci. 569, 259-268. 2. Akatsu T., Takahashi N., Debari K., Morita I., Murota S., Nagata N., Takatani 0. and Suda T. (1989) Prostaglandins NSC45,1--n
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 21. 28.
29. 30. 31 32 33
34. 35. 36. 37. 38.
promote osteoclast-like cell formation by a mechanism mvolving cyclic adinosine 3,5-monophosphate in mouse hon,~ marrow cell cultures. J. Bone Min. Res. 4, 29935. Assia E., Rosner M., Belkin M.. Solomon A. and Schwartz M. (1989) Temporal parameters of low energy iasei irradiation for optimal delay of post-traumatic degeneration of rat optic nerve. Bruin Res. 476, 205.212. Bhakoo K. K., Crockard H. A.. Lascelles P. C. and Avery S. F. (1984) Prostaglandin synthesis and oedema formation during reperfusion following experimental brain ischaemia in the gerbil. Stroke 15, 891.895. Bhattacherjee P. (1974) Autoradiographic localization of intravitreally or intracamerally injected (‘H) prosraglandin. Expl Eye Res. 18, 181-188. Bito L. Z. and Salvador E. V. (1972) Intraocular fluid dynamics: III. The site and mechanisms of prostaglandins transfer across the blood intraocular fluid barriers. Exp/ Eye Res. 14, 233-241. Bitterman P. B., Wewers M. D., Rennard S. I., Adelberg S. and Crystal R. G. (1986) Modulation of alveolar macrophages-derived fibroblast proliferation by alternative macrophage mediators. J. clin. Inuest. 77, 700~708. Burke W., Cotte L. J., Garvey J., Kumarasinghe R. and Kyniacau C. (1986) Selective degeneration of optic nerve fibers in the cat produced by pressure block. J. Physiol. 376, 461476. Dempsey R. J., Roy M. W., Meyer K., Cowen D. E. and Tai H. H. (1986) Development of cyclooxygenase and lipoxygenase metabolites of arachidonic acid after transient cerebral ischemia. J. Neurosurg. 64, 118-124. Feist R. M., Kline L. B., Morris R. E., Witherspoon C. D. and Michelson M. A. (1987) Recovery of vision after presumed direct optic nerve injury. Ophthalamology 94, I567- 1569. Feldman S. T., Rodriguez P., Cabrera A. C. and Mitchell M. (1990) The effect of temperature and trephination on eicosanoid release by donor corneas. Inoest. Ophlhalm. vis. Sci. 31, 486. Fitzpatrick L. R., Gaginella T. B., Haddox M. K. and Johnson L. R. (1988) Prostaglandin-mediated trophic effects on the rat duodenum: the role of polyamines. Proc. Sot. exp. Biol. Med. 189, 201-205. Glazer J. S. (1988) Topical diagnosis: prechiasmal visual pathways. In Clinical Ophthalmology (eds Duane T. D. and Jaeger E. A.), Vol. 2, pp. 65-66. J. B. Lippincott, PA. Hallenbeck J. M., Jacobs T. P. and Faden A. 1. (1983) Combined PGE,, indomethacin and heparin improves neurological recovery after spinal &hernia in cats. J. Neurosurg. 58, 7499754. Hanamura T., Shigeno T.. Asano T., Mima T. and Takakura K. (1989) Prostaglandin profiles in relation to local circulatory changes following focal cerebral ischemia in cats. Stroke 20, 803-808. Hotter G., Roseho-Catafu J.. Bulbena 0.. Gomez G., Colomer J., Saenz F. P. A., Cruz L. F. and Gelpi E. (1990) Prostaglandin E, and thromboxane Bz levels in rats subjected to pancreas transplantation. Prostaglandins 39, 5360. Hunt A. H. and Minckler D. S. (1983) Optic nerve axonal transport: basic aspects. In Biomedical Foundations of Ophthalmology (eds Duane T. D. and Jaeger E. A.), Vol. 1, Chap. 26. Harper and Row, PA. Hgu C. Y., Liu T. H., Yu J., Hogan E. L., Chao J., Sun G., Tai H. H., Beckman J. S. and Freman B. A. (1989) Arachidonic acid and its metabolites in cerebral ischemia. Ann. N. Y. Acud. Sci. 559, 282-295. Ianotti F., Crockard A., Ladds G. and Symon L. (1981) Are prostaglandins involved in experimental ischemic edema in gerbils. Stroke 12, 301-306. Johnson G. S. and Pastan I. (1971) Changes in growth and morphology of fibroblasts by prostaglandins. J. nam. Cancer Res. 47, 1357-1364. Jonsson H. T. and Daniel1 H. B. (1976) Altered levels of PGF in cat spinal cord tissue following traumatic injury. Prostaglandins 11, 5 l-6 1. Joseph M. P.. Lessel S., Rizzo J. and Momose J. (1990) Extracranial optic nerve decompression for traumatic optic neuropathy. Arch. Ophrhal. 108, 1091~-1093. Katz B., Sofonio M., Lyden P. D. and Mitchell M. D. (1988) Prostaglandin concentrations in cerebrospinal fluid of rabbits under normal and ischemic conditions. Stroke 19, 349-351. Leslie J. B. and Watkin W. D. (1985) Eicosanoids in the central nervous system. J. Neurosurg. 63, 6599668. Libby P., Warner S. J. C. and Friedman G. B. (1988) Interleukin I: a mitogen for human vascular smooth muscle cells that induces the release of growth-inhibitory prostanoids. J. clin. Invest. 81, 487498. Logn J. L., Lee S. M., Benson B. and Michael U. F. (1986) Inhibition of compensatory renal growth of indomethacin. Prostaglandins 31, 2533261. Marouf L. and Azar D. T. (1985) Reversible visual loss following optic nerve injury. Ann. Ophfhal. 17, 5822584. Ment L. R., Stewart W. B., Duncan C. C., Cole J. S. and Pitt B. R. (1985) Beagle puppy model of perinatal cerebral infarction. Acute changes in regional cerebral prostaglandins during hemorrhagic hypotension. J. Neurosurg. 63, 899-904. Ment L. R., Stewart W. B., Duncan C. C. and Pitt B. R. (1986) Beagle puppy model of perinatal cerebral insults. Cerebral blood flow changes and intraventricular hemorrhage evoked by hypoxemia. J. Neurosurg. 65, 847.-850. Ment L. R., Stewart W. B., Duncan C. C., Pitt B. R. and Cole J. S. (1986) Beagle puppy model of perinatal cerebral infarction. Regional cerebral prostagiandins changes during acute hypoxemia. J. Neurosurg. 65, 851-855. Ment L. R., Stewart W. B., Duncan C. C., Pitt B. R. and Cole J. S. (1987) Beagle pup model of brain injury: regional cerebral blood flow and cerebral prostaglandins. J. Neurosurg. 67, 2788283. Miller N. R. (1990) The management of traumatic optic neuropathy. Arch. Ophthal. 108, 10861087. Naveh N., Rosner M., Zborowsky-Gutman L., Rosen N. and Weissman C. (1990) A comparison of Argon and Neodynium: YAG laser iridotomy effect on prostaglandins E, and blood aqueous barrier disruption. Ophrhal. Res. 22, 253-259. Naveh N. and Weissman C. (1990) The correlation between excessive vitreal protein levels and prostaglandins PGE, levels on the blood retinal barrier. Prostaglundins 39, 147-156. Peyman G. A., Bennet T. D. and Velchek i (1975) The effects of intravitreal prostaglandin on retinal vasculature. Ann. Ophthal. 7, 7299134. PreudHomme Y., Demolle D. and Boeynaems J. M. (1985) Metabolism of arachidonic acid in rabbit iris and retina. Invest. Ophthal. vis. Sci. 26, 13361342. Salmon J. A. and Higgs G. A. (1987) Prostaglandins and leukotrienes as inflammatory mediators. Br. med. Bull. 43, 2855296. Samuelson B. (1987) An elucidation of the arachidonic cascade. Drugs 33, Suppl. 1, 2-9.
Prostaglandin
E, changes in the retina and optic nerve of the eye with injured optic nerve
225
39. Seiff S. R. (1990) High dose corticosteroids for treatment of vision loss due to indirect injury to the optic nerve. Ophthal. Surg. 21, 389-395. 40. Shohami E., Shapira Y., Sidi A. and Cotev S. (1987) Head injury induces increased prostaglandin synthesis in rat brain. J. cerebr. Blood Flow Metab. 7, 5863. 41. Shwartx M. (1987) Molecular and cellular aspects of nerve regeneration. CRC Biochem. 22, 89-110. 42. Shyne-Athwal S., Chakraborty G., Gage E. and Ingoglia N. A. (1988) Comparison of posttranslational protein modification by amino acid addition after crush injury to sciatic and optic nerves of rats. Expi Neurol. 99, 281-295. 43. Siminoff R. and Bito L. Z. (1982) The effects of prostaglandins and arachidonic acid on the electroretinogram: evidence for functional cyclooxygenase activity in the retina. Curr. Eye Res. 1, 635-642. 44. Sokal R. R. and RohIf F. J. (1969) Biometry, the principles and practice of statistics in biological research. W. H.
Freeman, San Francisco. 45. Stoll G., Trapp B. D. and Griffin J. W. (1989) Macrophage function during Wallerian degeneration of rat optic nerve: clearance of degenerating myelin and Ia expression. J. Neurosci. 9, 2327-2335. 46. Weissmann G. (1987) Pathogenesis of inflammation. Effects of the pharmacological manipulation of arachidonic acid metabolism on the cytological response to inflammatory stimuli. Drugs 33, 28-37. 47. White F. V., Toews A. D., Goodman J. F., Novicki D. L., Bouldin T. W. and Morel1 P. (1989) Lipid metabolism during early stages of Wallerian degeneration in the rat sciatic nerve. J. Neurochem. 52, 108>1092. 48. Wiley M. H., Feingold K. R., Grunfeld C., Quensey-Hunbeus V. and Wu J. M. (1983) Evidence for CAMP-independent inhibition of S-phase DNA synthesis by prostaglandins. J. biol. Chem. 258, 491496. (Accepted
7 May
1991)
