Jin Ma, Ting Chen,2, Yiwei Wang, Chan Zhao, Donghui Li,Meng Wang, Linyang Gan, Yong Zhong*

1Department of Ophthalmology, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College,Beijing 100730, China 2Department of Ophthalmology, Renmin Hospital of Wuhan University, Wuhan, 430060 China

Key words: nonarteritic anterior ischemic optic neuropathy; rat model; posterior optic nerve head; blood flow; in vivo

NONARIERIIIC anterior ischemic optic neuropathy (NAION) is one of the major causes of acute optic neuropathy in the elderly.1,2Despite the wide consensus on the“ischemic” nature and the gross anatomical location of this disease as the name implies, the exact underlying hemodynamic mechanism within the optic nerve head(ONH) remain to be fully elucidated.

Ihe ONH is vascularized by two major arterial systems：3the superficial nerve fiber layer is mainly supplied by retinal arterioles； the deep major part of the ONH, subdivided into prelaminar, lamina and retrolaminar regions, are principally supplied by direct branches of the posterior ciliary arteries (PCAs). It has been suggested that it is the deep ONH regions,supplied by PCAs, that probably show more decrease in circulation4and are involved in the pathogenetic changes in NAION.5-7However, most current available techniques for bloodflow measurement in NAION patients, such as Laser Doppler Flowmetry (LDF), laser speckleflowgraphy, can only measure the bloodflow of superficial nervefiber layer in NAION patients.3,8LDF is one of the non-invasive techniques for assessment of volumetric bloodflow.9It has been widely used to measure the hemodynamics of cerebral vascular after stroke,10,11and also in the measurement of bloodflow in the optic nerve head12and subfoveal choroid.13Rodents are significant animal model in physiology and therapeutics, and similar LDF technique has been developed to measure the vascular bed of the rat eyes.14

A rat NAION (rNAION) model with similar funduscopic and pathological changes to those of human NAION wasfirst established by Bernstein in 2003.15,16Ihe key event to generate this model is to induce photochemical thrombosis of the microvasculature in the ONH by direct irradiating the ONH with argon green laser after intravenous injection of Rose Bengal(RB), which results in ischemic ONH edema.17With the rNAION model, some studies have observed the deeper ONH circulation alterations in vitro ONH tissues.15,18,19Ihe strength of these results, however, is limited by the “in vitro” nature of the experimental approaches.In addition, the microcirculation status of ONH was only measured at the very early phase (no more than 3 days) after disease induction, which only provided a snapshot of the dynamic process and may obscure the speed and duration of arterial bloodflow reduction in ONH during disease progression.

In our previous studies, the structure and function changes have been longitudinally monitored.15,19Io shed new light on the characteristics of the hemodynamic changes in rNAION at different time points, dynamic changes in bloodflow kinetics of the ONH were measured “in vivo”.

MATERIALS AND METHODS

Animals

All protocols were approved by the Experimental Animals Committee and were consistent with the Declaration of Helsinki or the NIH statement for the Use of Animals in Ophthalmic and Vision Research. Male pathogen-free Sprague-Dawley (SD) rats (weighing 200-220 g, kept at 20-22 °C) were used for this study.Each rat in this study underwent a complete ophthalmic examination to rule out ocular abnormalities.

Induction of rNAION

Ihe experimental protocol for rNAION induction has been reported previously.15,19In brief, 1 ml/kg Rose Bengal (RB) ［2.5 mM in phosphate-buffered saline(PBS), 90% purity； Sigma-Aldrich, St. Louis, MO, USA］was injected via femoral vein and the right ONHs were immediately photo-activated using an argon-laser photo-coagulator (Ultima 2000 SE Argon, Coherent inc.,Santa Clara, CA, USA) with the following parameters：wave length 532 nm； power 50 mW； laser spot size 500 um； duration 12 seconds. Ihe posterior segments of the eyes were examined immediately after induction to rule out hemorrhage. For the laser group, the ONH of the right eyes were illuminated with the laser without intraperitoneal injection of RB； and vice versa for the RB group (only injected with RB via the femoral vein). Ihe naive group received no intervention as the name indicates.

Fundus photography and fundus fluorescein angiography

Fundus photography and fundus fluorescein angiography (FFA) were performed after anesthetization and mydriasis using a slip lamp equipped with a camera(ASA 800, incandescent； Nikon, Iokyo, Japan) and Spectralis HRA+OCI (Heidelberg Engineering, Heidelberg, Germany), respectively. For FFA, angiographic images centered on the ONH were taken at intervals of approximately 1 second immediately after injection offluorescein sodium for about 20 seconds, and continued at intervals of about 10 seconds for up to 5-9 minutes.Fundus photographs were taken in naive rats (n=2)and rNAION 3 h, 1 d, 5 d, 14 d and 90 d after disease induction (n=3 for each time point). FFAs were performed for naive rats (n=2) and rNAION (n=3) in 3 hours after disease induction. Ihe early phase of FFA was defined as before 1.5 minutes after injection, and late phase was defined as over 3 minutes after injection.

Histologic preparation

Animals were euthanized after deep anesthetization,and the eyes were removed immediately and thenfixed in 4% paraformaldehyde-phosphate-buffered saline (PFPBS) for 1 h. Iissues were then embedded in paraffin,sectioned, and stained with hematoxylin-eosin (H&E)or toluidine blue for histologic examination. Histologic examinations were performed in naive rats (n=2), laser-only rats and RB-only rats (1 d after treatment, n=2 for each group), and rNAION rats 1 d and 90 d after disease induction (n=2 at each time point).

Laser dopplerflowmetry

Longitudinal hemodynamic changes of posterior ONH were monitored with laser doppler flowmetry(LDF)device (moorVMS-LDF2, Moor Instruments Ltd., UK)on day 3, day 7, day 14, day 21 and day 40 after disease induction in rNAION rats (n=3 at each time point)and in naive rats. Ihe surface laser Doppler probe(VP3.0, Moor Instruments Ltd., UK), with a tip diameter of 0.50 mm, operated at the wavelength of 785 nm. As previously described,20the optic nerve of rat was exposed through a lateral canthus incision. Ihe LDF probe was placed perpendicular to the surface of the optic nerve and as close as possible to the eyeball.Ihe volume of LDF measurement was approximately 1 mm3. Ihe ONH bloodflow kinetics, measured in perfusion units (PU) on a real-time basis, was monitored for about 3 minutes till the readout reached steady state. PU is the basic index of LDF measurement and is the Doppler shift value produced by RBCflowing. It is a relative unit to measure the microcirculatory blood flow in local deep tissue and the change of PU value directly shows the changes of microcirculatory bloodflow of tissue.

Statistical analysis

All data were reported as mean ± SD. Statistical analysis was performed using SAS (version 9.3, SAS Institute Inc, Cary, NC) to evaluate longitudinal changes in ONH blood perfusion over time. Statistical differences were evaluated by one-way ANOVA, Student’s t-test and Bonferroni adjustment for multiple comparisons. A two-sided P value less than 0.05 was considered statistically significant.

RESULTS

Fundus photography and FFA of rNAION

Fundus photography (Fig. 1) showed that, in contrast to fundus of naive eyes, the edema of ONH in the rNAION eyes was observed at 3 hour after disease induction and reached maximum on day 1 after disease induction. On day 5, profound resolution of ONH edema was observed in the rNAION eyes, and the appearance of ONH returned to normal on day 14； on day 90,however, the ONH appeared to be pale and reduction in size.

FFA (Fig. 2) revealed the normalfluorescein perfusion of the choroidal and retinal vasculature in the early phase and normalfluorescein distribution of fundus in the late phase in naive rats (n=2). In rNAION rats 3 hours after disease induction, however, filling defects in the choroid and ONH were observed in the early phase, and marked dye leakage from the ONH was present in the late phase.

Histopathology

As compared with naive eyes, no abnormal changes were observed in the retina and ONH in the laser-only rats 1 day after laser treatment and RB-only rats 1 day after RB injection. In the rNAION rats, however, obvious ONH edema, usually associated with peripapillary retinal detachment, was observed on the day 1 after disease induction； and on the day 90, a reduction in number of retinal ganglion cell (RGC) axons, as well as tissuefibrosis and cellular infiltration were observed,while cell densities of inner nuclear layer (INL) and outer nuclear layer (ONL) remained unchanged (Fig. 3).

Figure 1. Fundus photographs of optic nerve head (ONH) in naive rats and rNAION rats. A. Ihe border of the ONH in a naive rat eye was distinct (arrow)； B. in rNAION rats, blurring of the ONH border (arrow) observed 3 hours after disease induction, indicating edema； C. diameter of the ONH enlarged and reached the maximum (arrow) 1 day after disease induction； D.five days after induction, edema of the ONH resolved almost completely； E. the appearance of the ONH returned to normal on day 14； F. the ONH was apparently pale with reduction in size on day 90. rNAION： rat model of nonarteritic anterior ischemic optic neuropathy； ONH： optic nerve head.

Figure 2. Representative FFA images of a naive eye and an rNAION eye 3 hours after disease induction. A. normal fluorescein perfusion of the choroidal and retinal vasculature in the early phase of FFA in a naive eye. B. normalfluorescein distribution in the fundus in the late phase of FFA in a naive eye； C. FFA imaging 3 seconds after injection (early phase) of an rNAION eye showedfilling defects in the choroid and the ONH (arrow)； D. late phase imaging of the same eye showed marked dye leakage (arrow)from the ONH. FFA： fundusfluorescein angiography.

Myelinated axon bundles of retinal ganglion cells(RGCs) were tightly packed and surrounded by thin pial septate in the optic nerve (ON) in naive rat eyes. On the day 90 after rNAION induction, however, there was a marked loss of central axons with increased septal thickness and shrinkage of the axonal bundles (Fig. 4).

Longitudinal measurement in hemodynamic changes of posterior ONH bloodflow in rNAION

Ihe mean bloodflow kinetics of naive eyes was 174.73±5.34 PU (n=3), and the mean bloodflow of rNAION eyes on day 3, 7, 14, 21 and 40 after NAION modeling were 116.12±8.93 PU (n=3), 94.89±1.83 PU (n=3),73.02±6.99 PU (n=3), 68.94±2.24 PU (n=3) and 60.92±4.58 PU (n=3) respectively, where a significant difference was noticed (F=175.06, P＜0.0001) with a rapid and pronounced drop of blood perfusion in posterior ONH observed (Fig. 5). Ihe reduction in blood perfusion was statistically significant at each time point after disease induction as compared with that in naive rats (all P＜0.0001). Ihe blood perfusion significantly decreased compared with that in naive rats, with a 33.55%, 45.69% and 58.21% reduction on day 3, day 7 and day 14, respectively. Ihe differences were significant between the data of day 3 and day 7 (t=4.66,P=0.008), day 7 and day 14 (t=4.81, P=0.006), day 3 and day 14 (t=9.47, P＜0.0001). No significant differences demonstrated between neither day 14 and day 21 (t=0.90, P=1.00), nor day 21 and day 40 (t=1.76,P=1.00).

Figure 3. Representative histologic sections of ONH and retina in naive and rNAION eyes (H&E stain). A. longitudinal sections of the ONH in naive eyes, B. in the laser-only eyes and C. in the RB-only eyes showed normal histology and anatomy of the optic nerve and peripapillary retina. D. On day 1 after rNAION induction, the ONH was edematous with thickened nervefiber bundles (double asterisks) and peripapillary retinal detachment (arrow). E. On day 90 after induction, there was a reduced number of RGC axons (long arrow) accompanied by gliosis and cellular infiltration (short arrow). F. naive eyes, G. laser-only eyes, H. RB-only eyes, and I. eyes 1 day after rNAION induction. Ihe RGCs in peripapillary retina were closely packed in a single layer with normal density. J. On day 90 after rNAION induction, there was an obvious reduction in density of the RGCs in retina, while the cell densities in the INL and ONL remained mainly unchanged. (magnifying power： A-E, 50×； F-J, 200×). RB： rose bengal； RGC： retinal ganglion cell； INL： inner nuclear layer； ONL： outer nuclear layer.

Figure 4. Representative cross sections of ON in naive and rNAION eyes (Ioluidine blue stain, 200×). A. cross section of ON in naive eye showed tightly packed axonal bundles by the pial septate (S)； B. cross section of ON on day 90 after rNAION induction, an apparent reduction in central axonal bundle density with increased septal thickness was observed.AxB： axonal bundle. ON： optic nerve.

DISCUSSION

Nowadays, plenty of experimental researches using rNAION have been conducted to investigate cellular inflammation,21oligodendrocyte death22and neuroprotective effects of drugs.23,24Ihe bloodflow of ONH has been observed in vitro in rNAION. Bernstein15used intra-cardiac perfusion of India ink and found a severe reduction in the filling of ONH capillaries 30 minutes after induction. Chuman et al. also found the reduction of capillary perfusion in the ONH 3 days after induction by using the same method.19However, Nicholson et al. used perfusion of fluorescent marker (FIIC-BSA)and showed a quite different kinetic of microcirculation impairment in the posterior segment of ONH with detectable but minimal loss of capillary perfusion 4 hours after disease induction and a much more significant loss of perfusion 1 day after induction.18It is worth noting that in these studies, microcirculation status of the ONH was only measured at the very early time after disease induction.

Figure 5. Ihe hemodynamic changes in blood flow kinetics of posterior ONH in naive rats and rNAION rats at different time points after disease induction. Ihe posterior ONH blood perfusion measurements in rNAION rats were significantly lower than that in naive rats (all P＜0.0001).Statistically significant reductions of optic nerve bloodflow were detected between day 3 and day 7, day 7 and day 14 after disease induction (both P＜0.01).

In this study, we produced the rNAION following the method of Bernstein.9Optic disc edema appeared as a hallmark in the early phase of disease and the FFA revealed early hypofluorescence followed by late hyperfluorescence in the optic disc. Histopathological study showed tissue edema followed by loss of axonal bundles, thickening of septa and gliosis in the optic nerve. Ihese results demonstrated that there was a selective damage to capillaries in ONH, followed by ONH edema, andfinally optic atrophy with loss of RGC.Iherefore, we believed that the rNAION in our study was successfully established as Bernstein did.

Ihen, we longitudinally measured the dynamics of bloodflow kinetic changes in ONH of rNAION by placing LDF probe directly 1 mm anterior to the optic nerve(the posterior ONH), where the most prominent degenerative changes were detected histologically.25Interestingly, in rNAION, the ONH blood perfusion did not reduce uniformly in depth. A sharp reduction of blood perfusion was found in 14 days after the disease induction, while in the longitudinal observation of the posterior ONH bloodflow, the blood perfusion remained low steadily afterwards, indicating that the ischemic status was then stable. Ihe underlying mechanism may lie in that in the early time, the capillary is blocked byfibrin,capillary leakage and extracellular edema result in intracellular edema and/or capillary compassion,18while in the late time, the impaired blood flow autoregulated25,26after photochemical thrombosis of microvasculature in the ONH.

In rNAION, we found that as the progress of the disease, even when the edema of the ONH had resolved, the density of RGCs still reduced. Ourfinding was consistent with the report in 2016.27Ihe coupling association between bloodflow and neural activity in neural tissues has been demonstrated in literatures.28Relationship had been also found between ischemic status and deteriorated visual function, damaged structure and other morphometric changes.29,30Ihus we hypothesized that the reduced bloodflow might be the reflection of the reduced metabolic demand associated with neuronal loss and connective tissue changes.

Despite we firstly observed in vivo the dynamic feature of posterior ONH bloodflow in rNAION in this study, limitations of this study should be addressed.Firstly, the approach used to measure the posterior ONH bloodflow was invasive and the surgical exposure of the posterior ONH was of high skilled operation.Secondly, we did not measure RGCs density of the retina which might be associated with the bloodflow of posterior ONH, thus we were not able to provide a quantitative assessment of the neurodegeneration.Ihirdly, functional evaluations of the ON, e.g. visual evoked potential, were not carried out in this study.Ihese limitations should be kept in mind when interpreting the data and should be taken into consideration for studies in future.

In summary, we successfully established rNAION.With this model, our study characterized the in vivo dynamic feature of posterior ONH blood perfusion reduction at different time points. Our findings, to some extent,provide a reference for understanding the hemodynamic changes occurred in rNAION. Further studies on ameliorating the early bloodflow with rNAION will provide valuable information on the therapeutic approaches.

Conflict of interest statement

All authors declared no conflict of interests.