The anti-oxidative Effect of Chrysanthemum Extract on Retinal- light Damage of Mice

Background: Apoptosis of photoreceptor cells and oxidative stress of RPE in age-related macular degeneration (AMD) could be promoted by photopic oxidative stress. In our study we are aim to study the protective effect of chrysanthemum extract on light damaged retina of mice. Methods: In vitro, ARPE-19 cells are incubated and divided into four groups: the control, the light damaged, the low and high dose-chrysanthemum extract groups. The last three groups were dropped in zero, low and high concentration of chrysanthemum extract separately before exposing to light. Cellular viability and Reactive Oxygen Species(ROS) production were measured by MTT and immunouorescence. In vivo, C57BL/6J mice were divided into four groups as above mentioned. Low and high concentration of chrysanthemum extract were given by continuous intragastric administration before being exposed to white light. Retinal function was evaluated by electroretinogram. Optical coherence tomography and Fluorescein fundus angiography were used to observe the morphology and vessels. HE staining and TUNEL immunouorescence for presenting morphology and apoptosis of isolated retina. Results: Viability of ARPE-19 cells decreased and ROS production increased after the light damaged. However, treatment with chrysanthemum extract, viability improved and ROS declined. After light injury, dysfunctional retina, destroyed morphology and increased apoptosis rate were observed in mice especially in RPE and photoreceptor layer. Treatment with chrysanthemum extract, retina function improved as well as structure of RPE and photoreceptor layers. Rate of apoptosis decreased via the raised concentration of anti-oxidative enzyme superoxide dismutase (SOD), catalase (CAT) and glutathione peroxidase (GSH-Px). Conclusions: Preventive chrysanthemum oxidative-stress induced light which indicating Chrysanthemum have potential of preventive for results show that LD group are much thinner than the control group of ONL from -600 to 600μm to the chrysanthemum extract light injury, the LC and HC group are much improved, especially the -600 to -300μm to 600μm optic with statistically signicant. Results through analyzing the INL thickness illustrated that only in the peri-retina have outstandingly improved


Introduction
Age-related degeneration (AMD) has been a common disease in developed countries and has a tendency to occur in more young people as the aging population raising in the world. According to a study of meta-analysis, the number of patients with AMD could reach to 0.2 billion in 2020 and add to 0.3 billion until 2040 [1]. AMD is characterized by the aggregation of drusen and the destroy of RPE layer in the early period, develop to wet type with neovascular or dry type with geographic atrophy (GA) in the late gradually [2,3]. The mechanism of AMD is not completely clear now, some studies show that the inducing factor could be relative with exposing to high intensity of light and lacking of anti-oxidative food in daily diet, which both lead to ROS produce in excess [4][5][6]. Chrysanthemum considered as a common herbal in Chinese Medicine for thousands of years, chrysanthemum contains anti-oxidative materials especially the micro molecules of avonoids, which are useful in anti-oxidative stress caused by a variety of color lights and scavenging oxygen free radicals [7][8][9]. The main purpose of this study is to observe the protective effect of chrysanthemum extract on the lighted injured RPE cells and retinal of mice. Discuss the possible mechanism of anti-oxidative stress.

Extraction of Chrysanthemum:
The chrysanthemum of experiment is produced in Hangzhou, China, and the petals were used to prepare for water extraction. After soaking chrysanthemum petals in water that 8 times weight of materials and boiling for 0.5 hour. Keep the liquid and residue was boiled for another time. Repeat the process for 3 times and then combine all the liquid obtained. Concentrate the collection to low and high concentration of extract. (1g chrysanthemum extract obtained from 4.255g materials) [10]. The content of total avonoids in chrysanthemum was 5.0%, according to Technical speci cation for inspection and evaluation of health food.

Cell experiment:
2.2.1. Cell culture: ARPE-19 cells (China Center for Type Culture Collection) were cultured in DMEM / F12 medium (Hyclone, the USA) containing 10% fetal bovine serum, 100U/mL penicillin and 100U mg/mL streptomycin, and incubated in 5% CO2 under constant temperature of 37 ℃ and humidity. When the cells grow exponentially, it will be used in the follow-up experiment.

The toxicity of chrysanthemum extract to ARPE-19 cells:
The safe concentration and toxicity of chrysanthemum extract to ARPE-19 cells were detected by MTT. AREP-19 cells approximate number of 10 ×103 cells were inoculated into 96-well plates with 100 μL per well. When the cells grew exponentially, the concentrations were 0.2mg/mL, 0.4mg/mL, 0.8mg/mL, 1.0mg/mL, 1.2mg/mL, 1.5 mg/mL, 2.00mg/mL and 5.00mg/mL respectively. of chrysanthemum extract were added to 96-well plate separately and cultured for 24 hours. After the end of culture, 20μL MTT (5mg/mL) solution was added to each well and cultured in CO2 incubator for 4 hours. MTT was discarded and 150μL DMSO solution was added. After being fully dissolved, the cell activity was evaluated by measuring absorbance (A value) at wavelength of 492 nm. According to the results, the safe concentration was applied for the follow-up experiment.

Light induced damage of RPE cell:
After the cells were digested by trypsin, the cells were made into cell suspension and inoculated in the culture plate. The cells were irradiated directly with LED white light with the intensity of (2500 ±500) Lux for 24 hours, after the cells grew exponentially [11] Totally 32 mice of C57BL/6J of male with the age of 6 weeks were obtained from the SPF Biotechnology (Beijing, Co,.Ltd), the mice were raised in 12/12h light and dark with the temperature of 23+2℃, relative humidity of 55% in the Tianjin Eye Institute, The food and water could take freely. All the procedures were coherence with international standard of laboratory animal use and care and in accordance with the principles and guidelines of the Chinese Council Animal Care, the experiment was approved by the Institutional Animal Care and Use Committee of Tianjin Medical University. Mice were experimented or sacri ced by inhaling the iso urane, all the steps to minimize suffering.

The groups division and chrysanthemum extract intervention:
Mice were randomly divided into control group (n = 8), light damaged group (LD, n = 8), low-dose chrysanthemum group (LC, n = 8) and high-dose chrysanthemum group (HC, n = 8). After concentrated and dried by water extraction, the extracts were crushed into powder and suspended with 0.9% saline when it was given to intragastric administration. In the range of effective concentration, 0.23g/kg/d for the low dose group, and 0.38g/kg/d for the high dose. 0.2ml/ per day for 8 weeks. The dose concentration was also similar with the study by Dong Lumeng et al. about the protective effect of chrysanthemum on Parkinson's mice [12].

Light damage method:
One day after intragastric administration, the light injury mode were as follows: light damage group(LD), low and high dose of chrysanthemum group (LC and HC)were treated with continuous white light of 9500±500lux for 7 days, 4 h /d after pupil dilated , accumulating 28 hours in total, each mouse was separated by transparent box during light exposing in order to avoid crowding. The function and morphology of retina were examined 7 days after photic injury.

Electroretinogram (ERG):
12 hours after dark adaptation, mice were anesthetized by inhaling 2% iso urane and xed on the animal experiment table. Dilate both of eyes and certain gel to keep the cornea transparent. Connecting reference electrode to the head, annular corneal electrode to the eyeball, and the ground electrode to the tail of mouse. Visual electrophysiological apparatus (Roland Germany) was used to record the amplitude of wave a, b and Ops under the stimulating light of 0.01cds/m 2 and 3.0 cds/m 2 in dark. After 10 minutes of adaptation of photopic, amplitude of a and b under 3.0 cds/m 2 was recorded.

Optical Coherence Tomography (OCT):
Mouse was lied prone on the animal experimental platform after being anesthetized by inhaling 2% iso urane, 0.5% tropicamide phenylephrine eye drops (Santen) were used to dilate pupils and carbomer eye gel keeping the cornea transparent. Phoenix eye testing equipment for animals (Phoenix research labs, model: Micron IV) was used to scan the retina in vivo and gain morphological imaging of each layer.

Fluorescein fundus angiography (FFA):
2% sodium uorescein was injected intraperitoneally immediately after nish the OCT examination. The state of lling of the retinal vessels at arterial phase, venous phase and arteriovenous phase was observed at the same gain value. Application of angio-tool 0.6a that a software download free was to analyze the relative parameters consist of vessel area, vessel area percentage and the total number of junctions.

HE staining and TUNEL:
The mice were sacri ced 3 days later by Inhaling iso urane, ophthalmic vessels blood collected, and the eyeballs were taken out at the same time. 12 hours after formalin xation of the eyeball, part of cornea was cut off and xed in fresh formalin for another 8 hours. After dehydration and para n liquid immersing at 40℃ for 6 hours, eyeballs were completely embedded Para n block which would be sliced into 4μm of thickness. HE and Tunel immuno uorescence staining (Tunel uorescence kit provide by Dalian meiliun biotechnology Co.,LTD.)were performed to analyze the apoptosis rate of each retinal layer for the all groups.

Anti-oxidative enzyme assay:
Superoxide dismutase (SOD), Catalase (CAT) and Glutathione (GSH-Px ) assay kit: Support by Nanjing Jiancheng Bioengineering Institute. Collected blood from ophthalmic artery was centrifugated at 4℃, 3000r/min for 10 minutes in order to obtain the serum. Keeping them at -80℃ to measure the enzyme activation of SOD, CAT and GSH-Px.

Statistical methods:
All the parameters including the amplitude of a and b wave in the examination of ERG, the vessel area, the vessel percentage and the total number of junctions measured and analyzed in FFA and angio-tool, the apoptosis rate in each layer of retina are presented in the term of χ̅ ± s , one-way ANOVA of SPSS22.0 software was to evaluate the parameter. P<0.05 has been considered as statistically signi cant. The results of MTT assay showed that the cell viability of the control group was 99.98% as the reference. The scores of cell viability were 97. 64%, 98. 52%, 96. 18%, 96. 24%, 94. 42%, 90.65%, at 0.2 mg/mL, 0.4 mg/mL, 0.8 mg/mL, 1.0mg/mL, 1.2mg/mL, 1.5mg/mL respectively. When the concentration was 2.00 mg/mL, chrysanthemum extract could signi cantly inhibit the viability of ARPE-19 cells and the cell viability was decreased to 88. 15% of the normal (P<0. 05) signi cantly and decreased to 64.66% of the normal (P<0.001) if the concentration was 5 mg/mL. Therefore, the concentration of chrysanthemum extract was selected no more than 1.2 mg/mL for the follow-up experiment. The LC groups with 0.4mg/mL and HC with 1.0mg/mL.

Animal examinations:
3.2.1. The retina changes in LD group: OCT and HE staining: As the imaging of OCT show, when the retinal structure was observed in vivo, it could be seen clearly of each layer (Figure 2A), the signal of high re ection that shaping of arched had been found between the RPE and interdigitation zone (IZ) layer in LD group, which also make the outer segments of photoreceptors (OSP) ,the ellipsoid zone (EZ) and myoid zone (MZ) layer to be deformed. (Figure 2B). At mean while the HE staining expressed the same structure and change with OCT imaging. The obviously damaged presented in OCT and HE staining that were similar with the AMD of early phase (Figure2.3). Imaging and Comparing each group in different zone (Figure4), the thickness we further analyze of inner nuclear layer (INL) and outer nuclear layer (ONL) of retina, the results show that LD group are much thinner than the control group of ONL from -600 to 600μm distance to the optic nerve, however, treated with chrysanthemum extract before light injury, the LC and HC group are much improved, especially from the -600 to -300μm and from 300 to 600μm distance to the optic nerve with statistically signi cant. Results through analyzing the INL thickness illustrated that only in the peri-retina have outstandingly improved (Figure5, Table1).

The Outcomes of ERG:
The wave of a, b and Ops shaped in Figure 6, under the 0.01cds/m 2 dark reaction in LD group, the amplitude of b-wave in LD group was 41.6% lower than control group, and the amplitude of b-wave in the LC and HC groups was 83.5% and 120.6% higher than LD group under the stimulate of 0.01cds/m 2 in the scotopic reaction (P < 0.05 in LC , P < 0.01 in HC. Under 3.0 cds/m2 dark reaction, the a-wave amplitude in LD group was 22% lower than control group, in LC group increased by 50.8% and HC and HC 118.5%. Compared with LD group, the amplitude of a wave in HC group was signi cantly higher than LD group (P<0.05) in the stimulation of 3.0cds/m 2 in photopic reaction (Figure 7). The data with amplitude and implicit time are displayed in Table 2. 3.2.3. The vessel change measured by FFA: The central vascular with the range of diameter between 2 to 4mm (Figure8), the LD group increased by 35% compared with the control group, while the LC and HC groups decreased by 20.8% and 15.7% respectively. In the total number of junctions, LD group increased by 64.4% than the control group, while the LD and HD groups decreased by 34.2% and 23.3% than the LD respectively. There was no statistical difference among each group, however there is a signi cant change between the LD and Control groups according to t -test of independent samples (P<0.05) as presented in Figure 9.

The apoptosis rate of retina measured by TUNEL assay:
There was outstanding increase of apoptosis cells in the LD group however much fewer in LC and HC groups (Figure10). The quantity of apoptotic cells counted by image J in LD group increased statistically signi cant. The apoptosis rate of RGC, INL and ONL layers were calculated separately ( Figure 11). The outcomes show signi cant differences of RGC, INL and total apoptosis rate among the four groups. Apoptosis rate of RGC cells in LD was 9.04% higher than the control group. LC group was 73.18% lower than LD group(P<0.01), HC group was 57.28%(P<0.01).

The enzyme activation of SOD, CAT and GSH-Px:
In the mouse fed up with continuous intragastric administration of chrysanthemum extract, the enzyme activity of SOD was increased with the mean value of 75.50±3.2U/ml and 75.57±6.33 U/ml in the LC and HC group respectively. CAT and GSH-Px show a more certain tendency of activity with the mean value of 1048.61±85.18 U/ml (P<0.01) and 1005.56±65.73 U/ml (P<0.01) in LC and HC groups separately. As depicted in gure 12, the CAT in HC group and the GSH-Px in LC and HC group are even more activating than Control group but with no statistically signi cant(Table3).

The relative between ROS and AMD:
The light damage as a method to establish model of oxidative stress of mice for more than 40 years, which could lead to photoreceptor destroyed [13]. Acute or chronic way is applied commonly by researchers, however studies by Chulbul M et al have shown that the generation of AMD is related to subacute in ammation [14][15][16]. Therefore, the use of strong light irradiation for a moderate time may be helpful for modeling. As we know that too powerful or longtime of light would lead to retinal phototoxic reaction and produce quantity of reactive oxygen species (ROS) that considered as the main metabolite of oxidative stress and attribute to oxidative damage reaction [17,18]. Recent study in clinical found that IS/OS is less detailed to divide the retina layer before.
The new more detailed zone named myoid zone (MZ) and ellipsoid zone (EZ), especially the EZ illustrated are corelative with the inner segment ellipsoid (ISE) that consist of large intensity of mitochondria [19]. When too much ROS can't be solved by mitochondria, aging or destroyed RPE cells accumulate some substances of lysosome non-degradable and then lipofuscin formed. These substances include dysfunctional mitochondria and the outer segments of swallowed photoreceptor cells [18]. The deposition of lipofuscin in the early stage does not cause abnormal cell metabolism. However, on account of RPE cells act as permanent cells in the body, and the accumulated lipofuscin cannot be diluted by proliferation. Long-term involvement of intracellular lipofuscin affecting the function of mitochondria and lysosomes even the proteolysis system [20,21]. The decreased proteolytic capacity leads to the accumulation of extracellular oxidized proteins, which resulting in drusen [22] and may cause the apoptosis raised in retina. At present, some studies have found that abnormal autophagy and DNA damage response (DNA damage response, DDR) is an important mechanism of AMD, and the main reason is also related to the increase of ROS level [23]. Therefore, oxidative damage is not only an important factor in cell senescence, but also forms a malignant pathological circulation mechanism of AMD. In our study we observed the damaged of EZ, OSP, and the RPE cell layer from OCT imaging in vivo and HE staining in vitro. All above we could come to conclude that process of anti-oxidative is important for prevent the retina from AMD.

The effect of chrysanthemum extract on oxidative stressed retina
Among the remarkable achievement of Traditional Chinese medicine in the study of anti-oxidant damage, chrysanthemum is a plant of asteraceae specie, natured as sweet, bitter and slightly cold for the body and attribute to the lung and liver meridians. They have good function to dissipate heat, protect eye from decreasing vision and fatigue. Chrysanthemum extract is rich in avonoids and polysaccharides In the experiment, we found that the electrophysiological function indicated by ERG and the morphology of retina displayed by OCT and HE staining in vivo and in vitro were signi cantly improved after the intervention of chrysanthemum extract. This is similar to the study of cell by Kim, I. S, which consistent with the conclusion that chrysanthemum can improve the loss of neuronal vitality induced by MPP+, reduce the apoptosis rate and increase the expression of Bcl-2 [27]. Suyao sun et al also found that wild chrysanthemum extract can reduce the level of ROS in cells and has the effect of anti-ultraviolet damage and skin photoaging in vitro [28]. On the opposite, result of FFA, the micro vessels diameter 2-4 mm analyzed shows that the vessel area, the vessel area percentage and the total number of junctions between the vascular are largely increased in the LD group compared with the control group. However, LC and HC groups have no signi cantly improved.
So, we further measure apoptosis in our study, in LD group, the apoptosis of retinal ganglion cells, inner nuclear layer cells and outer nuclear layer cells increased, especially the rst two, and the apoptosis rate of ganglion cell layer decreased signi cantly in LC and HC mice. In biochemical of blood test, the activation of antioxidant enzyme of CAT and GSH-Px increased signi cantly after intervention with chrysanthemum extract. The CAT and the GSH-Px were even better than the control group. However, the SOD increased but with no statistically signi cant which may have the relationship with the number of samples.
Therefore, we can speculate that the functional and morphological protection of chrysanthemum on retina may play a role by antagonizing oxidative stress and reducing the pathway of mitochondrial apoptosis. The mechanism of the protective function needs more research.

Conclusion
In our study, preventive administration of chrysanthemum extract improves both function and morphology of retina injured by light exposing and reduces the oxidative-stress by increasing SOD, CAT and GSH-Px, which indicating Chrysanthemum have a potential of preventive measure for AMD.

3.Competing interests:
The author declare that they have no competing interests.   Tables   Due to technical limitations, table 1, table 2 and table 3    OCT and Fundus photos of every groups. OCT: Compared with the LD group, the structure of LC and HC retina kept relative integrity and had no signals of high re ection in RPE layer. The red arrow showed the change position caused by light injury.

Figure 5
The thickness of INL and ONL. There were signi cant thinner in LD group when analyze the thickness on the zone of peri-retina from 300 to 600μm distance to optic nerve of INL. The tendency become more obviously from -300 to -600μm and from 300 to 600μm distance from the optic nerve. The detailed data could be seen in table1.

Figure 6
The shape of full-eld ERG waves in scotopic and photopic. Figure5 legend: The rst trough considered as a wave, the rst peak considered as b wave, the amplitude of b was valued from the trough point to the peak. Ops was measured under the stimulation of 3.0cds/m2, and lter out from 20 to 40Hz. The ruler of the amplitude and implicit time was 12.5 v and 5ms respectively.  FFA of the central fundus in the artery-venous period. Figure7 legend: Take the optic disk as the center position, imaging of FFA recorded at the arteriovenous phase, intensity of micro vessels between 2-4mm diameters were obvious increased in LD group. The ampli ed detail circled red showed the vessel area percentage and the total number of junctions.

Figure 9
Comparation of micro-vessels analyzed by the angio-tool software. After processing the data by angio-tool, the parameter collected and analyzing by SPSS, there were no signi cant statistically between the four groups by ANOVA, but testing by t-test of independent samples, the LD group had a remarkably rising compared with the control group. (*P<0.05). A: vessel area, B: vessel area percentage, C: total number of junctions.  to analyze the apoptosis rate of RGC(A), INL(B), ONL(C) and total number(D) of each group. Apoptotic cells of RGC in LC and HC groups were much lower than the LD groups(**P<0.01), so did the INL(**P<0.01). There was no signi cant change in ONL of apoptosis rate. The total number of apoptosis rate has decreased largely in LC and HC groups(**P<0.01).

Figure 12
The anti-oxidative enzyme of SOD, CAT, GSH-Px Figure11 legend: SOD(A), CAT(B) and GSH-Px(C), LC and HC group had a quite large increase in the enzyme activation of SOD, CAT and GSH-Px, and even more activating than the control group of last two.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. OnlineTABLE1.Png OnlineTABLE2.Png OnlineTABLE3.Png