CAMK2D and Complement Factor I–Involved Calcium/Calmodulin Signaling Modulates Sodium Iodate-Induced Mouse Retinal Degeneration

Abstract

Purpose: To investigate the effect of Ca²⁺/calmodulin-dependent protein kinase II (CAMKII) δ subtypes (CAMK2D) on sodium iodate (NaIO₃)-induced retinal degeneration in mice.

Methods: Bioinformatics analysis and Western blot experiments were used to screen the significantly differentially expressed genes in age-related macular degeneration (AMD) disease. CAMK2D knockdown and overexpression models were constructed by lentivirus (LV) infection of adult retinal pigment epithelial cell line-19 (ARPE-19) cells in vitro. Flow cytometry was used to detect ARPE-19 cell apoptosis induced by NaIO₃. In vivo, CAMK2D knockdown and overexpression mouse models were generated by infecting mouse retinal pigment epithelium (RPE) with adeno-associated virus (AAV). Retinography, optical coherence tomography (OCT), and histological analysis (hematoxylin and eosin staining) were used to detect NaIO₃-induced retinal structural changes in mice. Electroretinography (ERG) was used to detect NaIO₃-induced retinal function changes in mice. TdT-mediated dUTP nick-end labeling (TUNEL) staining was used to detect the apoptosis of retinal cells induced by NaIO₃. RNA sequencing (RNA-Seq) and bioinformatics analysis were used to screen for target genes affected by CAMK2D in CAMK2D-overexpressing ARPE-19 cells. And flow cytometry, OCT, and ERG were used to evaluate the regulatory effect of CAMK2D on target genes.

Results: Bioinformatics analysis found the expression of genes related to Ca²⁺ signal was significantly reduced in AMD patients. Western blot showed that in a mouse model of dry AMD induced by NaIO₃, CAMK2D expression in RPE-Choroid tissue significantly lower than normal mice. In vitro, our results showed that overexpression of CAMK2D in ARPE-19 cells decreased apoptosis induced by NaIO₃ and knockdown increased apoptosis. In vivo, CAMK2D overexpression in RPE cells can attenuate the retina degeneration induced by NaIO₃ and CAMK2D knockdown aggravated degeneration. The bioinformatics analysis indicated that CAMK2D might affect AMD pathology through complement factor I (CFI). In vitro, knockdown of CFI in ARPE-19 cells increased apoptosis induced by NaIO₃. In knockdown CFI ARPE-19 cells, overexpression of CAMK2D reduced the above apoptosis. In mice retina, CFI knockdown can aggravate the retina degeneration induced by NaIO₃. In knockdown CFI mice, overexpression of CAMK2D in RPE can attenuate the above retina degeneration. Western blot confirmed that CAMK2D regulated the expression of CFI in mice.

Conclusions: CAMK2D can attenuate the retinal degeneration induced by NaIO₃, which was achieved by regulating the CFI.

Age-related macular degeneration (AMD) is a leading cause of visual dysfunction worldwide, as a progressive retinal degenerative disease influenced by both environmental and genetic factors.¹ Genetic factors play an important role in AMD, and it is estimated that the genetic factors can account for 45% to 70% of the disease.² Retinal pigment epithelium (RPE) dysfunction is considered to be the key cause of AMD pathogenesis. For now, no methods could cure AMD.^1,3 To explore the abnormal gene expression of AMD patients, we analyzed the gene expression profile GSE125564 from RPE cells of AMD patients in Gene Expression Omnibus database. The data indicated that the genes expression related to Calcium ion binding and Voltage gated calcium channel activity were significantly reduced in AMD patients compared with non-AMD patients. Calcium (Ca²⁺) regulates a range of cellular processes as a highly versatile intracellular signal.^4,5 Ca²⁺/calmodulin-dependent protein kinase II (CAMKII) is a multifunctional kinase as a central coordinator and executor of Ca²⁺ signal transduction.^6,7 Studies have shown that CAMKII was involved in the regulation of multiple biological processes such as cell apoptosis^8,9and angiogenesis.^10–13 The occurrence and development of AMD are closely related to RPE cell apoptosis and angiogenesis.^14–16 However, the effect of CAMKII on RPE has not yet been studied. Our Western blot experiments show that in a mouse model of dry AMD induced by NaIO₃, CAMK2D expression in RPE-choroid tissue significantly lower than normal mice. Therefore we evaluated the effect of CAMK2D on dry AMD in vitro and in vivo, and explored its mechanism of action. 

Material and Methods

Bioinformatics Analysis

Data sets were downloaded from Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo). The microarray data of GSE125564 is based on the GPL23159 platform (Affymetrix Clariom S Assay, Human), which contains a dataset of systems-level transcriptome analysis of RPE samples differentiated from induced pluripotent stem cells (iPScs) of four advanced AMD and age-matched three non-AMD donors.¹⁷ The data sets had been normalized through log transformation. The DEGs were identified by the “limma” package through NetworkAnalyst (V.3.0) online database (https://www.networkanalyst.ca/). The volcano plot was graphed through NetworkAnalyst online database, and the GO functional enrichment analysis was graphed through “ggplot” package in R 4.3.0 software.¹⁸ 

Cell Culture

At 37°C temperature incubator with 5% CO₂, ARPE-19 cells were cultured in complete medium (89% DME/F12, 10% FBS, and 1% penicillin-streptomycin).¹⁹ 

Lentiviral Infection of ARPE-19 Cells

The lentivirus-based vectors were used to construct overexpression of CAMK2D and shRNA-mediated lentivirus vectors were used to construct knockdown of CAMK2D (Gene-Chem, Shanghai, China). Lentiviral infection effects were displayed in Supplementary Figure S1. Puromycin 2 µg/mL was used to selected ARPE-19 cells and then amplified (Beyotime Biotechnology, Shanghai, China). ShRNA-mediated lentivirus vectors were used to construct knockdown of complement factor I (CFI) (Gene-Chem). 200 µg/ml geneticin was used to selected ARPE-19 cells and then amplified (Beyotime Biotechnology). All the sequences are displayed in Supplementary Tables S1 through S3. 

Quantitative Real-Time Reverse Transcription Polymerase Chain Reaction (qRT-PCR)

Total RNA was extracted from the ARPE-19 cells using an PrimeScript RT-PCR kit (TaKaRa, Shiga, Japan). The polymerase chain reaction (PCR) reaction conditions were as follows: pre-degeneration at 95°C for 30 seconds, Reps: 1; degeneration at 95°C for five seconds and annealing/stretching at 60°C for 30 seconds, Reps: 40. The primers of CAMK2D: forward primer: 5′-CAGTGACACCTGAAGCCAAAGA-3′; reverse primer: 5′-GTGCATCATGGAGGCAACAGTA-3′. The endogenous control was β-Actin. The relative mRNA level of CAMK2D was calculated by the 2^−ΔΔCt method.²⁰ 

Western Blot Analysis

ARPE-19 cells were lysed using a mixture of RIPA, PMSF, and phosphatase inhibitors, and then total protein was extracted. Mice were anesthetized by intraperitoneal injection of excess pentobarbital sodium. Carefully separate the tissue around the eye, cut the optic nerve and quickly remove the eye, and then place it on ice. Carefully remove the conjunctiva, cornea, lens, and the rest to retain the complete eye cup. The retinal layer was carefully removed and the RPE-chorioscleral tissue was preserved. The pigmented tissue on the inner surface was scraped with a 1 mL syringe needle and collected into a centrifuge tube. The sclera and residual tissue were discarded. A sufficient amount of pigmented tissue was collected and spun in a centrifuge at 4°C. The dark brown precipitate was RPE-choroidal tissue. Total proteins were extracted from RPE-choroidal tissue using a Minute Total Protein Extraction Kit (Invent Biotechnologies, Eden Prairie, MN, USA) according to the manufacturer's instructions.²¹ The protein was configured into samples of concentration. Equal amounts of samples were subjected to electrophoresis and then transferred to the PVDF membrane.²² The endogenous control was β-actin. The intensities of the bands were measured by ImageJ software.²² 

Flow Cytometry

PE Annexin V Apoptosis Detection Kit with 7-AAD was purchased from Becton, Dickinson and Company.²³ FITC Annexin V Apoptosis Detection Kit with PI was purchased from Beyotime Biotechnology. According to the manufacturer's instructions, the ARPE-19 cells were labeled for apoptosis.²² The data was analyzed by FlowJo Software.^22,24 

Animals and Model of NaIO₃-Induced Retinal Degeneration

The adult Male MC57BL/6J mice aged 6–8 weeks were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China). Mice were maintained in 23° to 25°C, 12-hour light/dark cycle and free access of water and food.^25,26 All animal experiments were reviewed and approved by the Experimental Animal Ethics Committee of Jinzhou Medical University and conformed to the “Guide for the Care and Use of Laboratory Animals” published by the National Institutes of Health (Publication, 8th Edition, 2011). The study is reported in accordance with ARRIVE guidelines (https://arriveguidelines.org).²² Mice were intraperitoneally injected with 25 mg/kg NaIO₃ once. One week after injection, fundus photography and optical coherence tomography (OCT) of Micron IV retinal imaging camera system were performed. The model was successful when the retina showed deposits like drusen. 

Construction of AAV9 Vectors

AAV9 vectors were constructed by the Shanghai Genechem Co. Ltd. (Shanghai, China). AAV9-rpe65p-MCS-EGFP-3Flag-SV40 PolyA was the vector element sequence. Mouse CAMK2D(NM_001025438) was constructed into the AAV9 vector using the restriction enzymes NcoI/NcoI. Mouse CAMK2D (mir155[CAMK2D]) was constructed into the AAV9 vector using the restriction enzymes BsrGI/NheI. The specific RPE65 (retinoid isomerohydrolase RPE65) promoter was proved to be an RPE tissue-specific promoter. After AAV9 infects the retina, expression of the foreign gene under the regulation of this promoter normally occurs only in the RPE.²⁷ 

Intravitreal Injection of Viral Vector

Mice were anesthetized by the intraperitoneal injection of pentobarbital sodium (80 mg/kg). After mydriasis with compound tropicamide eye drops, a sterile disposable injection needle (33-gauge) was used to pierce the sclera gently about 1 mm behind the limbus under an operating microscope. Subsequently, a 33-gauge microsyringe was used to enter through the scleral puncture port. Under the microscope, the tip of the needle was located behind the lens and in front of the retina, and about 1 µL of virus was slowly injected. After the injection was completed, the eyeballs were smeared with gatifloxacin eye gel, and the mice were promptly rewarmed and revived. Gatifloxacin eye gel was applied daily for three days after surgery to prevent infection.²⁶ 

Retinography and OCT Imaging

Mice were anesthetized through the intraperitoneal injection of pentobarbital sodium (80 mg/kg). Their pupils were dilated with tropicamide and corneas were protected by gatifloxacin eye gel. Retinography and OCT were taken by Micron IV retinal imaging camera system (Phoenix Research Laboratories, Pleasanton, CA, USA).²² Based on OCT results, the peripheral nerve thickness of the optic nerve head was measured by Image-Pro Plus 6.0 software. 

Histological Analysis

The eyes were enucleated and embedded in paraffin. Retinal section tissues were sliced for hematoxylin and eosin (H&E) staining. Images were taken by a microscope.^22,28 

ERG

Mice were dark adapted overnight and then anesthetized, and their pupils were dilated.²⁹ The full-field ERGs were recorded by Micron IV retinal imaging camera system. The parameters were as follows: flash strength: 6.2 log (cd/m²), duration: 10 ms, aiming light: 6.2 log(cd/m²), 50 Hz clean filter, delay between sweeps: 5.0 sec. The amplitude of a-wave and b-wave were statistically analyzed separately.²² 

TUNEL Staining

Mice were anesthetized by intraperitoneal injection of excess pentobarbital sodium. Cardiac perfusion with normal saline solution was performed first and continued until the liver turned white. Then, normal saline solution was replaced by 4% paraformaldehyde to continue perfusion until the body stiffness of the mice. Carefully separate the tissue around the eye, cut the optic nerve and quickly remove the eye, and then place it in eye-specific fixative solution at 4 °C. After 72 hours, carefully remove the conjunctiva, cornea, lens, etc. to retain the complete eye cup. The eye cup was gradually dehydrated and embedded in optimal cutting temperature compound embedding agent (O.C.T. Sakura; Sakura Finetek USA, Inc., Torrance, CA, USA) and then frozen overnight. The next day, the eye cup was made into a cryosection with a thickness of 10 µm for TUNEL staining. Cryosections of eyeball were stained by the In Situ Cell Death Detection Kit (Roche, Basel, Switzerland) according to the instructions. Nuclei were stained with DAPI, and photographs were taken by fluorescence microscope (Leica, Wetzlar, Germany). The apoptosis cells were counted and averaged in five visual fields selected randomly.^30,31 

RNA Seq and Bioinformatics Analysis

RNA library sequencing was performed on the Illumina Hiseq 2500/4000 by Gene Denovo Biotechnology Co., Ltd (Guangzhou, China). Bioinformatic analysis was performed using Omicsmart, a real-time interactive online platform for data analysis (http://www.omicsmart.com).³² 

Statistical Analysis

Each experiment was repeated a minimum of three times. GraphPad Prism 8 (GraphPad, San Diego, CA, USA) was used for statistical analysis and graph. T-test was used for two groups statistical analyses. For three or more groups, statistical analyses were calculated with one-Way analysis of variance, followed by Tukey tests for between-group analysis. P < 0.05 was considered as a significant difference.²² 

Results

Bioinformatics analysis found that the Ca²⁺-related molecular functions (MF) were downregulated in AMD and Western Blot indicated that CAMK2D expression was downregulated in a mouse model of dry AMD. Eight hundred twenty-five significant differentially expressed genes (DEGs) of GSE125564 were identified, include 475 upregulated genes and 350 downregulated genes in AMD samples, as shown in the Volcano plot (P < 0.05, |log fold change (FC)|>1, Fig. 1A). The downregulated genes were used to perform gene ontology (GO) through the PANTHER classification system. Sorted by descending order of P value for the GO enrichment terms, the term of calcium ion binding ranked the first one and hit the largest number of genes in molecular functions. The term of voltage gated calcium channel activity ranked third in molecular functions (Fig. 1B). NaIO₃-induced retinal degeneration is an animal model for AMD and frequently used to study cellular and molecular mechanisms underlying AMD. We constructed the model, and our Western blot showed that the CAMK2D expression in the NaIO₃ group was significantly lower than the normal group (P < 0.05, Figs. 1C, 1D). This suggests that CAMK2D expression was downregulated in the dry AMD model induced by NaIO₃. 

Figure 1.

 

CAMK2D expression in the dry AMD. (A) Volcano plot showing 825 significant DEGs of GSE125564 (Adjusted P value < 0.05, │log FC│ > 1). (B) The molecular functions terms in GO of downregulated genes by PANTHER. (C) CAMK2D protein level in RPE-choroid tissue. (D) Quantification based on (C) (n = 3, **P < 0.01).

Knockdown or Overexpression of CAMK2D Increased or Decreased ARPE-19 Cell Apoptosis Induced by NaIO₃ In Vitro

As assayed by the PE Annexin V Apoptosis Detection Kit with 7-AAD, treatment with 1250 µg/mL NaIO₃ for 24 hours, total apoptosis rate in LV-CON313 + NaIO₃ group was approximately 10%. With NaIO₃ treatment, the total apoptosis rate in RNAi (119303) + NaIO₃ group was significantly increased compared with the LV-CON313 + NaIO₃ group. Without NaIO₃ treatment, the knockdown of CAMK2D did not affect ARPE-19 cells apoptosis (Figs. 2A, 2B). With NaIO₃ treatment or not, the total apoptosis rate in LV-CAMK2D was lower than LV-CON335 (Figs. 2C, 2D). The data demonstrated that knockdown of CAMK2D in the ARPE-19 cells (RNAi [119303]) increased cells apoptosis and overexpression (LV-CAMK2D) decreased cells apoptosis induced by NaIO₃. 

Figure 2.

 

Flow cytometry analysis of apoptosis of ARPE‐19 cells. (A) About knockdown of CAMK2D, the percentage of cells in each quadrant is presented. (B) The total apoptosis rate in four groups were statistically analyzed based on (A). (C) About overexpression of CAMK2D, the percentage of cells in each quadrant is presented. (D) The total apoptosis rate in four groups were statistically analyzed based on (C). (Q2 represents late apoptosis, Q3 represents early apoptosis, Q2 + Q3 represents total apoptosis. n = 3 per group, ns: P > 0.05, **P < 0.01, ****P < 0.0001).

Overexpression or Knockdown of CAMK2D Did Not Alter the Retinal Structure in Normal Mice

The above results showed that CAMK2D overexpression in ARPE-19 cells decreased cells apoptosis and knockdown increased cells apoptosis from oxidative damage. We further performed adeno-associated virus 9 (AAV9)-CAMK2D overexpression vector (AAV9-CAMK2D) and AAV9-CAMK2D knockdown vector (AAV9-shRNA-CAMK2D) in RPE cells of C57BL/6J mice to clarify the role of CAMK2D. Three weeks after the intravitreal injection of AAV9 vector, infection efficiency and CAMK2D expression levels were measured. We selected mouse retinas successfully expressing green fluorescent protein for Western blotting through Micron IV retinal imaging camera system (Fig. 3A). Western blotting indicated that CAMK2D was increased in RPE after AAV9-CAMK2D infection and downregulated in AAV9-shRNA-CAMK2D (Figs. 3B–E). To determine whether CAMK2D can change retina, fundus camera, OCT, and H&E staining were used to visualize retinal structures (Figs. 3F–H), and ERG was used to examine retinal function (Figs. 3I–K). AAV9-CAMK2D and AAV9-shRNA-CAMK2D showed no significant differences compared with the normal and control. These results indicate that neither overexpression nor knockdown of CAMK2D in RPE affects the retinal structure and function in normal mice. 

Figure 3.

 

Overexpression or knockdown of CAMK2D did not alter the retinal structure and function in normal mice. (A) Retinas expressing green fluorescent protein (GFP) were measured through Micron IV retinal imaging camera system. (B) Western blotting of CAMK2D after the injection of AAV9-CAMK2D vector. (C) Quantification based on (B). (D) Western blotting of CAMK2D after the injection of AAV9-shRNA-CAMK2D vector. (E) Quantification based on (D). (n = 3; *P < 0.05, **P < 0.01). (F) Fundus camera showed no significant changes in retinal structure after the AAV9 injection. (G) OCT showed no significant changes in retinal structure after the AAV9 injection. (H) H&E staining showed no significant changes in retinal structure after the AAV9 injection. (I–K) ERG showed no significant differences of a-wave and b-wave after AAV9 vector injection. (n = 5; ns: P > 0.05).

CAMK2D Overexpression in RPE Cells Attenuated the Retinal Degeneration and Knockdown Exacerbated Retinal Degeneration Induced by Oxidative Stress

In order to assess whether CAMK2D changes could affect the retina in oxidative damage, the mice were intraperitoneally injected with 25 mg/kg NaIO₃, and one week later the retina was examined through Micron IV retinal imaging camera system. Compared to the AAV9-CON427 + NaIO₃ group, less yellow-white exudation, thicker retina, and lighter structural damage in AAV9-CAMK2D+NaIO₃ (Figs. 4A–E). And ERG indicated that a-wave and b-wave amplitudes were higher in AAV9-CAMK2D+NaIO₃ than AAV9-CON427 + NaIO₃ (Figs. 4F–H). TUNEL-positive signals were lower in AAV9-CAMK2D + NaIO₃ group than AAV9-CON427 + NaIO₃ (Figs. 4I–J). On the contrary, thinner retina and more severe structural damage were in AAV9-shRNA-CAMK2D + NaIO₃ than AAV9-shRNA-NC + NaIO₃ (Figs. 4A–E). Moreover, lower a/b-wave amplitude (Figs. 4F–H) and higher TUNEL-positive signals (Figs. 4I–J) was in AAV9-shRNA-CAMK2D + NaIO₃ than AAV9-shRNA-NC + NaIO₃. 

Figure 4.

 

Overexpression of CAMK2D in RPE attenuated NaIO₃-induced retinal degeneration and knockdown exacerbated NaIO₃-induced retinal degeneration. (A) Fundus camera (white arrows indicate foci of yellow-white exudation). (B) OCT (black arrow indicates the OCT imaging location; red arrows indicate bulging deposits like drusen. Scale bar: 100 µm.). (C) H&E staining (green arrows indicate bulging deposits. Scale bar: 50 µm.). (D) OCT detecting retinal thickness (black circular arrow indicates the OCT imaging location. Scale bar: 100 µm.). (E) The retinal thickness quantitative analysis based on (D). (n = 5; ****P < 0.0001). (F) ERG results. (G) The a-wave quantitative analysis based on (F). (H) The b-wave quantitative analysis based on (F) (n = 5; **P < 0.01, ***P < 0.001, ****P < 0.0001). (I) TUNEL results indicated retina apoptotic cells (white arrows indicate apoptotic cells. Scale bar: 50 µm. Below is the organizational structure of partial enlargement of the green rectangle in above.). (J) Quantitative analysis the number of apoptotic cells based on of (I) (n = 5; **P < 0.01, ****P < 0.0001).

CAMK2D Regulates the Expression of CFI in ARPE-19 Cells Induced by NaIO₃

To explore the mechanism of CAMK2D, the ARPE-19 cells after up-regulation of CAMK2D expression was detected by RNA-seq. RNA-seq identified 163 DEGs, which include 62 upregulated transcripts and 101 downregulated transcripts after CAMK2D overexpression (Fig. 5A). Seventy-seven disease genes of AMD were found in online mendelian inheritance in man. Venn diagrams show that CFI as the only overlapping gene was found between 163 DEGs and 77 AMD disease genes (Fig. 5B). CFI has two transcripts, encoding CFI and novel protein, respectively. In RNA-seq, CFI was upregulated, and novel protein was downregulated. Western blot results showed that CFI in overexpression CAMK2D was upregulated and knockdown was downregulated after ARPE-19 Cells induced by NaIO₃ (Figs. 5C–F). 

Figure 5.

 

(A) DEGs of overexpression CAMK2D in ARPE-19 cells were shown in a volcano plot (P < 0.05, |log2FC| > log2(1.8)). (B) Venn diagrams show overlapping genes between disease genes of AMD and the DEGs. (C) Western blotting of CFI in ARPE-19 cells of overexpression CAMK2D. (D) Quantification based on (C). (E) Western blotting of CFI in ARPE-19 cells of knockdown CAMK2D. (F) Quantification based on (E) (n = 3 per group. ns: P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001).

CAMK2D Regulates CFI to Affect Retinal Degeneration Induced by NaIO₃

Western blotting indicated that CFI was downregulated in ARPE-19 cells after LV-shRNA-CFI (CFI[RNAi]) infection (Figs. 6A, 6B). As assayed by the Annexin V-FITC Apoptosis Detection Kit with PI, without NaIO₃ treatment, the knockdown of CFI did not affect ARPE-19 cell apoptosis. With NaIO₃ treatment, the total apoptosis rate in CFI(RNAi) + NaIO₃ group was significantly increased compared with the control (Normal + NaIO₃ and LV-CON389 + NaIO₃) group (Figs. 6C, 6E). The above data demonstrated that knockdown of CFI in the ARPE-19 cells increased apoptosis induced by NaIO₃. As assayed by the PE Annexin V Apoptosis Detection Kit with 7AAD, with NaIO₃ treatment, the total apoptosis rate in CFI(RNAi) + LV-CAMK2D + NaIO₃ group was significantly decreased compared with the control (CFI(RNAi) + LV-CON335 + NaIO₃ group) group (Figs. 6D, 6F). The results indicated overexpression of CAMK2D reduced the apoptosis induced by NaIO₃ in ARPE-19 cells of CFI knockdown. Next, we further investigated the effect of overexpression of CAMK2D in mice retina of CFI knockdown on NaIO₃-induced retinal degeneration. Western blotting indicated that CFI was downregulated in RPE-choroidal tissue after LV-shRNA-CFI (CFI[RNAi]) infection. CFI expression was higher than the Normal group after AAV9-CAMK2D infection. And CFI expression in the CFI(RNAi) + AAV-CAMK2D group was higher than the CFI(RNAi) group (Figs. 6G, 6H). The results indicated that overexpression of CAMK2D in RPE-Choroidal tissue can increase CFI expression. OCT examination of the mice retina revealed that, compared to the Normal + NaIO₃ group, the retina was thinner in CFI(RNAi) + NaIO₃ group. And the retina was thicker in CFI(RNAi) + AAV9-CAMK2D + NaIO₃ group than the CFI(RNAi) + NaIO₃ group (Figs. 6I, 6L). ERG indicated that a-wave and b-wave amplitudes were lower in CFI(RNAi) + NaIO₃ group than the Normal + NaIO₃ group. And the a-wave and b-wave amplitudes were higher in CFI(RNAi) + AAV9-CAMK2D + NaIO₃ group than the CFI(RNAi) + NaIO₃ group (Figs. 6J, 6K, 6M). The above results demonstrate that overexpression of CAMK2D can attenuate retinal degeneration induced by NaIO₃ in mice retina of CFI knockdown. 

Figure 6.

 

(A) Western blotting of CFI in ARPE-19 cells of CFI knockdown. (B). Quantification based on (A) (n = 3. ns: P > 0.05, **P < 0.01). (C) The total apoptosis rates were statistically analyzed based on (E) the Annexin V-FITC Apoptosis Detection Kit with PI. (D) The total apoptosis rates were statistically analyzed based on (F) the Annexin V-PE Apoptosis Detection Kit with 7AAD. (E) The percentage of cells in each quadrant is presented on the Annexin V-FITC Apoptosis Detection Kit with PI. (F) The percentage of cells in each quadrant is presented on the Annexin V-PE Apoptosis Detection Kit with 7AAD (n = 3. ns: P > 0.05, **P < 0.01, ****P < 0.0001). (G) Western blotting of CFI in RPE-Choroidal tissue of mice. (H) Quantification based on (G) (n = 3. ns: P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001). (I) The retinal thickness quantitative analysis based on (L). (n = 5; **P < 0.01, ****P < 0.0001). (J) The a-wave quantitative analysis based on (M). (K) The b-wave quantitative analysis based on (M) (n = 5; *P < 0.05, **P < 0.01, ****P < 0.0001). (L) OCT detecting retinal thickness (Scale bar: 100 µm.). (M) ERG results.

Discussion

AMD as a neurodegenerative disease is closely related to aging and genetics.¹ The RPE plays a key role on maintaining retinal homeostasis through multiple biological functions.²⁷ We found that the expression of genes related to Ca²⁺ signal was significantly reduced in AMD patients through bioinformatics analysis. CAMKII is a multifunctional kinase as a central coordinator and executor of Ca²⁺ signal transduction,^6,7 which was involved in cell apoptosis^8,9 and angiogenesis.^10–13 These biological processes are closely related to AMD.^14–16 However, there was a lack of direct evidence to elucidate the role of CAMKII in AMD. Our Western Blot experiments show that in a mouse model of dry AMD induced by NaIO₃, CAMKII δ subtypes (CAMK2D) expression in RPE-Choroid tissue significantly lower than normal mice. Accordingly, we hypothesized that CAMK2D may be involved in regulating the pathological process of AMD. And then this hypothesis was verified by experiments. 

In vitro, we found that CAMK2D was involved in apoptosis induced by NaIO₃ in ARPE-19 cells. Knockdown increased ARPE-19 cells apoptosis and overexpression decreased apoptosis, which was consistent with the conclusions of most studies about CAMKII in other cells^4,33–35 For retinal ganglion cells (RGCs) in stressful environment, basal levels of CAMKII can constantly afford them protection.^33,34 Overexpression of CAMKIIαB by pIRES-hyg3/αB enhanced RGC-5 cells survival against glutamate treatment.³³ The Paeoniflorin upregulated the level of p-CAMKII and then triggered AMPK activation against apoptosis induced by oxidative stress in ARPE-19 cells.³⁵ Correspondingly, in RGCs with knocked down levels of CAMKIIalphaB by specific siRNA, cell apoptosis increase induced by glutamate.³³ To further confirm the role of CAMK2D on apoptosis in vivo, gene transfer experiments were performed. The results showed that CAMK2D overexpression in RPE cells attenuated the retinal degeneration and knockdown exacerbated retinal degeneration induced by NaIO₃. Such results further demonstrate the important role of CAMK2D in retina, which may affect AMD developing. 

To find the mechanism through which CAMK2D affects AMD, RNA-seq experiments were conducted. We overlapped 163 DEGs and 77 disease genes of AMD to find the key genes which CAMK2D may regulate. CFI as a major factor in the complement system was the only overlapping gene. CFI, a complement inhibitor, regulates the complement system activation by degrading complement component 3b (C3b).³⁶ A growing body of research has shown that the complement pathway was engaged in the pathogenesis of AMD.³⁷ In aging, complement may act through affecting inflammation, metabolism, apoptosis and other biological processes.^38,39 Hyperactivation of the complement replacement pathway is one of the major causes of AMD.^38,40 Complement component 3 (C3) as the central point of the three-cascade activation pathway of the complement system can lead to apoptosis by forming the membrane attack complex via the complement-activated pathway.³⁸ CFI can reduce complement activity through restraining C3,¹⁹ which may reduce cells apoptosis. Moreover, haploinsufficiency of CFI is associated with decreased retinal thickness and is a strong risk factor for AMD development.^41,42 Therefore we considered CFI as a major signaling gene for validation. In RNA-seq, CFI was upregulated. Western blotting analysis confirmed that CFI in overexpression CAMK2D was upregulated and knockdown was downregulated after ARPE-19 Cells induced by NaIO₃. These suggest that CAMK2D might affect RPE cells apoptosis and AMD pathology through CFI. 

To further investigate whether CAMK2D could affect NaIO₃-induced retinal degeneration by regulating CFI, CFI knockdown and CAMK2D overexpression models were constructed In vitro and vivo. As assayed by the Flow cytometry, knockdown of CFI in the ARPE-19 cells increased apoptosis induced by NaIO₃ and overexpression of CAMK2D reduced the apoptosis induced by NaIO3 in ARPE-19 cells of CFI knockdown. Western blotting indicated that CAMK2D can regulate CFI expression. Overexpression CAMK2D in RPE-Choroid tissue can increase CFI expression. OCT examination revealed that CFI knockdown make retinal thickness thinning induced by NaIO₃ and overexpression of CAMK2D can increased retinal thickness in NaIO₃-induced mice of CFI knockdown. ERG indicated that CFI knockdown make retinal function more severe damage induced by NaIO₃ and overexpression of CAMK2D can alleviate this retinal dysfunction in NaIO₃-induced mice of CFI knockdown. The above results demonstrated that overexpression of CAMK2D can attenuate retinal degeneration induced by NaIO₃ in mice retina of CFI knockdown. 

In conclusion, our study indicates that CAMK2D can attenuate the retinal degeneration induced by NaIO₃, which may be achieved by regulating the CFI. But there still needs a lot of follow-up experiments to verify. Our results provide new insights into potential therapies for AMD. 

Acknowledgments

The authors thank Jianhua Huang for providing experimental technical guidance. 

Disclosure: W. Xu, None; L. Cao, None; H. Liu, None