Retinal Levels of Amyloid Beta Correlate with Cerebral Levels of Amyloid Beta in Young APPswe/PS1dE9 Transgenic Mice before Onset of Alzheimer's Disease

Objectives Retina abnormalities are related to cognitive disorders in patients with Alzheimer's disease (AD). Retinal amyloid beta (Aβ) can be labeled by curcumin. We measured Aβ content in the cerebrum and retina of APPswe/PS1dE9 (APP) transgenic mice with early age to investigate the correlation between cerebrum and retina. Methods APP mice and age-matched wild-type mice were investigated every month from age 2 months to 6 months to assess changes in Aβ content in the retina and cerebrum. At the beginning of each month, mice were fed a curcumin diet (50 mg/kg/day) for 7 consecutive days. The Aβ levels in the retina and cerebrum were measured by ELISAs. Correlations were identified between retinal and cerebral Aβ contents using Pearson's correlation. Results In the absence of curcumin, there was a significant correlation between Aβ contents in the retina and cerebrum of APP mice (r = 0.7291, P = 0.0014). With increasing age, Aβ-mediated degenerative change in the cerebrum (P < 0.001 in 5 months) and retina (P < 0.01 in 5 months) increased significantly. The inhibitory effect of curcumin on the Aβ level was significant in the cerebrum (P < 0.001) and retina (P < 0.01) of older APP mice in the early stage of life. Conclusion We observed a significant correlation between the Aβ content in the retina and Aβ content in the cerebrum of APP mice. Our data suggest an appropriate time to measure retinal Aβ. Although curcumin can label Aβ in the retina, it also suppresses Aβ levels and weakens the degree of correlation between Aβ in cerebrum and retina tissues.


Introduction
Early diagnosis of Alzheimer's disease (AD) is essential for treatment [1]. Most diagnostic methods for AD are based on clinical symptoms [2]. Commonly, people in the presymptomatic stage of AD have no clinical symptoms (including impairment in episodic memory).
"Amyloid beta" (Aβ), which denotes peptides of 36-43 amino acids, is a major pathologic hallmark in the central nervous system (CNS) of patients with AD [3,4]. Amyloidal precursor protein (APP) is first cleaved by the enzymes βsecretase and γ-secretase and then released into the space between cells [5]. Due to the different cleavage sites of γsecretase, Aβ lengths are different. Soluble Aβ oligomers are more toxic than deposited plaques [6]. Although Aβ plaques are found in the brains of many elderly people without AD, they might be denoting a presymptomatic stage of AD.
Auxiliary diagnoses involve the examination of cerebrospinal fluid and positron emission tomography (PET) of Aβ plaques [7,8]. The examination of cerebrospinal fluid Aβ is an invasive method. High cost of Aβ-PET limits its use in early diagnosis. Recent studies demonstrate the ability to detect CNS Aβ deposition via the use of plasma assessment of Aβ species [9,10]. Due to the blood-brain barrier, plasma amyloid beta levels cannot reflect the real condition in the brain. Plasma concentrations of Aβ40 and Aβ42 have been shown to increase with age and in early AD but may decrease with advancing AD. However, no significant differences in plasma Aβ concentrations have been reported between individuals with and without AD [11].
However, AD is a disease of the CNS, and Aβ may distribute in all parts of nervous tissue, including the cerebrum and retina [12]. Ocular amyloidal imaging has been used to diagnose AD and monitor, noninvasively, AD progression [13].
Although the identifiable difference in the retinal structure between AD patients and healthy people is not straightforward, and misdiagnoses can occur [14,15], retinal Aβ plaques which can be labeled with curcumin may improve diagnostic accuracy [16,17]. This method merits further study and development into a new diagnostic measurement. Moreover, curcumin is a safe, nontoxic lipophilic agent with antioxidant and anti-inflammatory properties [4]. Apart from being a valuable labeling agent, curcumin may also play an important part in the AD treatment without eliciting side effects [18,19].
The biological basis of curcumin-labeled examination of the retina is good coherence between the cerebrum and retina. Retinal Aβ plaques can reflect Aβ plaques in the cerebrum. When is the best time to early detect Aβ and diagnose AD through the retina? What is the consistency between the cerebral Aβ and the retinal Aβ? In this brief research report, we elucidated Aβ content and its coherence in the cerebrum and retina of mice with early-age before the onset of AD. 2.2. Animals. APPswe/PS1dE9 transgenic mice (APP) and age-matched wild type (WT) mice were provided by the Model Animal Research Center of Nanjing University (Nanjing, China). To exclude the effect of sex on results, only male mice were used. Animals were housed in cages in a room maintained at 22 ± 2°C and 60 ± 5% relative humidity under a 12 h light-dark cycle (lights on at 6 : 00 am). Water and food were freely available in their cages. Animal experiments were conducted outside of their housing area in a separate room.

Experimental Procedures.
According to previous studies, amyloid-beta plaques in CNS of AD mice are markedly formed at 6 months of age [20]. Curcumin was administered to 2-, 3-, 4-, 5-, and 6-month old APP and WT mice for 7 consecutive days using the intragastric (i.g.) administration route.
The dose of curcumin used in this study was chosen according to previous animal studies and clinical trials [16,21,22]. According to previous studies, at high dosages, curcumin might prevent short-term recognition but not spatial memory. No signs of neurogenesis were evident, but reduced neuroinflammation was observed. The dose of the intragastric (i.g.) administration route has been shown to reduce the risk of vascular inflammation in the brain of AD subjects [19,23].
Mice were sacrificed by neck amputation after the final administration. Then, the retina and cerebrums were isolated. Tissue samples were homogenized in RIPA Buffer (Beijing Solarbio Science & Technology, Beijing, China) at 1 : 10 (g/v) with 1% phenylmethylsulfonyl fluoride (Beijing Solarbio Science & Technology). Supernatant proteins were extracted after centrifugation (13,000 rpm or 20 min at 4°C). For each sample, 150 μL of extracted protein was used for detection. The Aβ concentration was quantified using a mouse total Aβ ELISA kit (Shanghai Yuanye BioTechnology, Shanghai, China) according to manufacturer protocols. 2.5. Enzyme-Linked Immune Sorbent Assay (ELISA). The Aβ level in the brain and retina was measured by ELISAs. Absorbance at 450 nm (at a reference wavelength of 690 nm) was measured by an absorbance reader (Sunrise™; Tecan, Geneva, Switzerland). The absorbance value was transformed into a concentration value by reading the absorbance of pure samples on a standard curve.
2.6. Aβ Immunohistochemistry. Briefly, after anesthetized, mice were perfused with saline until the limbs and the liver turn white, and then perfused with 4% paraformaldehyde until the tail became stiff. The brain tissue was dissected and incubated with 4% paraformaldehyde for 1 day. After being washed with PBS, the tissue was put in a centrifuge tube containing 30% sucrose solution until the brain tissue sunk to the bottom. A cryostat was used to cut the brain tissue into 25 μm thick brains. The sections were incubated in 1% BSA for 1 h and then incubated with β-amyloid antibody (1 : 500, Cell Signaling Technology) at 4°C overnight. The sections were washed 3 times with PBS, rinsed, and incubated with the secondary antibody at 37°C for 1 h. After staining with 4 ′ ,6-diamidino-2-phenylindole (DAPI) for 1 min, the sections were washed 3 times with PBS and imaged using a confocal fluorescence microscope.

Statistical Analyses.
Data are the mean ± standard error (SE). Prism v7.0 (GraphPad, San Diego, CA, USA) was employed for statistical analyses. Differences among multiple mean ± SE values were assessed by one-way and two-way ANOVA, followed by Bonferroni's post hoc test. Differences 2 Behavioural Neurology between two mean ± SE values were assessed by the unpaired t-test. The correlation between Aβ content in the retina and cerebrum in each group was tested using Pearson's correlation. P < 0:05 was considered significant. 3 Behavioural Neurology brain of APP transgenic mice older than 6 months [20]. Mice in the present study were aged 2-6 months, so Aβ plaques have not yet formed in the brain. Hence, we measured the Aβ concentration in the brain of each mouse using ELISAs. WT mice given or not given curcumin are represented as WT C and WT N , respectively. APP mice given or not given curcumin are represented as APP C and APP N , respectively.

Results
Results revealed that Aβ was present in all four groups (WT C , WT N , APP C , APP N ). Figure 1 shows the change in the Aβ level in the four groups with increasing age, as well as the effects of curcumin on the Aβ level in mouse brains. Aβ content was significantly higher in the APP N group compared with that in the WT N group (P < 0:05 from 3 month to 6 month). A two-way ANOVA (genotype × treatment) revealed a significant effect of genotype (Fð1, 8Þ = 100:5, P < 0:001) from 3 months to 6 months.
From 5 months of age, the difference in the Aβ level between the curcumin administration group and no curcumin administration increased. The inhibitory effect of curcumin on the Aβ level was significant in the brain in 5 months, with 38.82% suppression by curcumin (APP C vs. APP N , P < 0:001). Images of Figure 2 show a progressive increase in plaque load of APP mice. The Aβ plaques began to accumulate from 5 months of age in tissue level.

Retinal Content of Aβ in
Mice Aged 2-6 Months. Different from the cerebrum, there was no significant difference in the Aβ level in the retinas between the APP N group and the WT N group until 5 months in Figure 3. At 5 months of age, the difference between APP C and APP N was significant, with 15.96% inhibition by curcumin being recorded (P < 0:01).

Deposition Trend of Aβ in the Cerebrum and Retina.
There was an age-related increase in the Aβ content, as shown in Figures 1(f) and 2(f). Aβ accumulation in the brain was different from 2 months to 6 months of age. At 5 months of age, curcumin had a significant effect on the Aβ content. The retinal content of Aβ in mice aged 2-4 months was not sensitive to the effects of curcumin. Correlation analyses between the Aβ level in the retina and that in the cerebrum are shown in Figure 4. Using combined data from APP mice of all ages, the Aβ level in the retina was correlated positively with the Aβ level in the cerebrum. Without curcumin administration, the Aβ level in the retina was correlated significantly with the Aβ level in the cerebrum (r = 0:7291, P < 0:01). Coherence between the brain and retina was established gradually with increasing age.

Discussion
In 2009, Perez and coworkers suggested that Aβ deposition within the retina can contribute to retinal dysfunction [24]. Many studies have focused on finding the best time to detect Aβ deposition in the eye because it could aid the early diagnosis of AD in humans [25,26]. Aside from Aβ deposition, studies have focused on retinal function (e.g., light reflection) in young mice with AD [12]. However, Aβ-mediated degenerative change in the cerebrum and retina before the onset of AD has not been studied.
APPswe/PS1dE9 transgenic mice display early onset of Aβ deposition in the CNS [27,28]. Hence, APPswe/PS1dE9 transgenic mice could be an ideal model to study the Aβrelated pathogenic effects on the nervous system in earlystage AD and even stage before the onset of AD.
Using APPswe/PS1dE9 transgenic mice, our study reported that in early age just before AD onset, there was a correlation between amyloid-beta levels of cerebrum and retina. Researchers have found Aβ plaques on the retina with the aid of the fluorescence effect of curcumin [17]. We also investigated the effect of curcumin on Aβ levels in CNS. Oral administration of curcumin can not only label the amyloidbeta plaques in the retina which may be a window of AD noninvasive diagnosis as Koronyo et al.'s study but also suppress the amyloid-beta level and its correlation of cerebrum and retina.
Time is an important factor in AD pathology. Notably, the accumulation of amyloid-beta increased over time. At 5 months old, the Aβ content became significantly different between mice of APP and WT genotype. The difference between Aβ levels of curcumin and no curcumin groups was higher from 5 months to 6 months. Before 4 months, cerebral levels of amyloid-beta were more affected by genotype than curcumin.
In the cerebrum of young mice, the difference in the Aβ level between APP and WT came earlier than that in the retina. This finding may be because Aβ accumulates in the hippocampal region at the beginning of life and then diffuses to frontal and temporal cortices and other parts of the CNS [1]. With increasing age, the Aβ level in the retina of APPswe/PS1dE9 transgenic mice increased. This appears to resemble a process of age-related decrease in the number of synapses in retinal layers as a result of Aβ accumulation [29]. Aβ accumulation is the upstream event and can be  Behavioural Neurology indirectly reflected by age-related increase in presynaptic and a decrease in postsynaptic retinal proteins in retinal plexiform layers [3,29]. These phenotypes were similar to the brain. At this time (especially in 5-month-old mice), the amyloidal pathogenesis of AD compared with that in the normal retina could be distinguished. This may be the best time to detect amyloidal pathogenesis from retina for early diagnosis of AD. Several research teams have used curcumin to label Aβ in the retina and brain [30]. Curcumin can also suppress Aβ  accumulation. In our study, curcumin decreased the Aβ level in the brain and retina of APPswe/PS1dE9 transgenic mice in the early stage of life, but it had a more potent inhibitory effect in older mice. Feeding demethylcurcumin or bisdemethylcurcumin to APPswe/PS1dE9 double-transgenic mice can upregulate the NEP expression in the brain and reduce Aβ accumulation in the hippocampus and cortices of mice at 4.5-5.5 months of age [31].
In the absence of curcumin, the trends of Aβ accumulation in APPswe/PS1dE9 transgenic mice and WT mice were inconsistent and continued to increase with increasing age. However, in APPswe/PS1dE9 transgenic mice, there was a significant correlation between the Aβ content in the retina and brain. In addition to Aβ, hyperphosphorylation of tau proteins disrupts the retinal structure and may contribute to visual deficits seen in APPswe/PS1dE9 transgenic mice [24,32].
Besides the retina, other ocular changes occur in AD patients and animals: altered pupil flash response, Aβ aggregation in the lens, and abnormal pattern electroretinograms [33][34][35][36]. Hence, even though most AD-related disease occurs in the brain, AD can also affect the eye. The retina shares many features with the brain (embryological origin, anatomic (e.g., microvascular bed) and physiologic (e.g., bloodtissue barrier) characteristics) [14], so the relationship between the brain and retina merits further study.
Ocular imaging of Aβ in AD patients could facilitate noninvasive monitoring [13]. In humans, ophthalmic imag-ing methods are used to assess neurodegenerative disorders, such as AD and Parkinson's disease [15]. Presently and in the future, the relationship between the degree of cognitive impairment and retinal abnormality merits further the study (Table 1).
Aβ is also observed in virtually all people with Down syndrome aged > 40 years and leads to a clinical diagnosis of dementia [37]. Increased retinal nerve fiber layer (RNFL) thinning in adults with Down syndrome may represent early AD-related changes. Studies in other neurodegenerative diseases, such as multiple sclerosis and cognitive decline, have also suggested that a thinner RNFL may be a preclinical observation of dementia [38,39].
As a potential contrast agent of AD retinal diagnosis, pharmacokinetics of curcumin in wild type and APP mice should be further explored. A previous study investigated a magnetic resonance imaging contrast agent [40]. They reported that no significant differences were observed in the plasma or brain kinetics of wild type and APP mice. Similar to curcumin, this contrast agent was previously shown to cross the blood-brain barrier and bind to amyloid plaques in the brain of AD transgenic mouse.
With the advent of advanced imaging technologies and Aβ biomarkers for clinical use, it is now possible to identify the effects of Aβ accumulation through noninvasive imaging of ocular structures in live patients. Optical coherence tomography angiography (OCTA) was used in clinical trials  [42] * P < 0:05; * * P < 0:01; * * * P < 0:001; AD: Alzheimer's disease; MRI: magnetic resonance imaging; PET: positron emission tomography; CT: computed tomography; OR: odds ratio; AUC: area under the receiver operating characteristic curves (to assess the ability of RNFL thicknesses to discriminate AD cases from healthy people). 6 Behavioural Neurology of AD detection, but only revealed the structure of biological tissues, such as RNFL thickness and vessel density [39]. Therefore, a detection method based on pathological biomarkers is urgently needed. In the future, AD progression could be quantified by measurement of the retinal Aβ level. One limitation of the proposed method is that although curcumin can label Aβ in the retina, it also suppresses Aβ levels and weakens the degree of correlation between Aβ in cerebrum and retina tissues.

Conclusions
We observed a significant correlation between the Aβ content in the retina and Aβ content in the brain of young APP mice before the onset of AD. Our data provide a biological basis for supporting noninvasive detection of AD in the eye and also suggest a time to detect retinal Aβ. Although curcumin can label the Aβ, it can also suppress the Aβ level and weaken the degree of correlation.

Data Availability
All data are included in the manuscript. However, the raw data used and/or analyzed in the present study are available from the corresponding author on reasonable request.

Ethical Approval
The study protocol was approved by the Animal Care and Use Committee of the Medical School of Ningbo University (Ningbo, China). Animal experiments were undertaken according to the Guide for the Care and Use of Laboratory Animals (National Institutes of Health (Bethesda, MD, USA) publication number 80-23, revised 1996).