Background
Age-related macular degeneration (AMD) is a leading cause of severe visual impairment in the elderly, and the number of affected persons steadily increases as a consequence of demographic changes [
1]. Late stage AMD can be classified in a dry and a wet form, representing atrophic and neovascular processes, respectively. Wet AMD can be treated with intravitreal injections of anti-vascular endothelial growth factor (VEGF) medication, while no established treatment options exist for the dry form [
2]. The pathogenesis of AMD is characterized by the early presence of drusen, damage of the retinal pigment epithelium (RPE) and photoreceptor layer primarily affecting the macular region and a chronic inflammatory response in the retina leading to the expression of pro-inflammatory and pro-angiogenic factors [
2]. This immunological response in the retina can be regarded as para-inflammation and involves the reactivity of microglial cells [
3‐
5].
Microglial cells are the tissue macrophages of the central nervous system (CNS), including the retina. In the healthy adult retina, they are located in the plexiform layers from where they permanently scan the retinal environment with their motile protrusions [
6]. The expression of several receptors that are specific for the binding of chemokines, cytokines, complement factors, antibodies, or damage-associated molecular patterns enables these cells to recognize and immediately respond to pathological changes of their environment [
7‐
9]. Besides their supportive function in the healthy retina, microglia reactivity and age-related changes of microglia physiology contribute to degenerative pathologies of the retina and the entire CNS [
3,
4,
8,
10‐
17]. A large number of amoeboid shaped reactive microglia are detectable in the degenerating photoreceptor layer of AMD and retinitis pigmentosa retinas [
4]. These cells contain phagocytosed rhodopsin-positive particles [
4], and a recent report demonstrated that microglia do not only phagocytose dead cells but also take up living rods in a mouse model for retinitis pigmentosa [
13]. Therefore, microglial activation cannot be just regarded as a bystander effect but rather actively contributes to photoreceptor cell death during retinal degeneration.
Because of these findings, substances that modulate microglial reactivity such as minocycline are good candidates to prevent inflammation and dampen degenerative processes in the retina. Minocycline is a second-generation semi-synthetic tetracycline analog, which is used against Gram-positive and Gram-negative bacteria for over 30 years. Besides its bacteriostatic capacity, minocycline exerts anti-inflammatory, anti-apoptotic, and neuroprotective effects in experimental models of Parkinson’s disease [
18], multiple sclerosis [
19‐
22], neuropathic pain [
16], Alzheimer’s disease [
23,
24], and in a genetic mouse model of retinitis pigmentosa [
25].
In this study, we addressed the questions how minocycline modulates microglial reactivity in vitro and whether it protects from acute white light-induced retinal degeneration in the mouse. We selected white light exposure as it is an environmental risk factor that mimics several features of AMD in rodents including degeneration of photoreceptors and the retinal pigment epithelium [
26‐
28]. This model is also very useful for a temporal correlation of microglial responses with processes of retinal degeneration and RPE dysfunction [
29‐
31].
Methods
Reagents
Minocycline hydrochloride (M9511), Escherichia coli 0111:B4 lipopolysaccharide (LPS), and Z-Leu-Leu-Leu-al (MG-132) were purchased from Sigma-Aldrich (St. Louis, MO, USA).
Cell culture
BV-2 microglia were cultured in RPMI1640 with 5 % fetal calf serum (FCS) supplemented with 2 mM
l-glutamine, 1 % penicillin/streptomycin, and 195 nM β-mercaptoethanol at 37 °C in a humidified atmosphere of 5 % CO
2, as described previously [
32,
33]. BV-2 cells were pre-incubated for 30 min with 50 μg/ml minocycline or NaCl as vehicle control. Afterwards, the cells were stimulated with 50 ng/ml LPS, 50 μg/ml minocycline, or 50 ng/ml LPS plus 50 μg/ml minocycline for 24 h. After stimulation, cells or supernatant were harvested for further analysis. 661W photoreceptor-like cells were a kind gift from Prof. Muayyad Al-Ubaidi (Department of Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK, USA). 661W photoreceptor cells were cultured with Dulbecco’s modified Eagle’s medium (DMEM), high glucose with
l-glutamine, supplemented with 10 % FCS and 1 % penicillin/streptomycin. Cultures were maintained in a sterile humidified environment at 37 °C and 5 % CO
2 as described elsewhere [
34].
RNA isolation and reverse transcription
Total RNA was extracted from cultured BV-2 microglia or murine retinas using the NucleoSpin® RNA Mini Kit (Macherey-Nagel, Dueren, Germany). RNA was quantified spectrophotometrically with a NanoDrop 2000 (Thermo Scientific). First-strand complementary DNA (cDNA) synthesis was carried out with the Revert Aid H Minus First-strand cDNA Synthesis Kit (Fermentas, K1632).
Real-time RT-PCR
cDNA (25 ng) were amplified in a 10-μl reaction mixture consisting of 5 μl Fast Start Universal Probe Master (Rox) (Roche), 2 μl of primers (10 μM), 0.375 μl purified water, and 0.125 μl of dual-labeled Universal ProbeLibrary (UPL) probe (Roche Applied Science, Basel, Switzerland) with an Applied Biosystems 7900 HT Fast Real-time PCR system (Applied Biosystems, Carlsbad, CA, USA). The following reaction parameters were used: 10 min 95 °C hold, followed by 40 cycles of 15 s 95 °C melt, and 1 min 60 °C anneal/extension. Primer sequences and UPL probe numbers were as follows: CD68, forward primer 5′-ctctctaaggctacaggctgct-3′, reverse primer 5′-tcacggttgcaagagaaaca-3′, probe #27; activated microglia/macrophage whey acidic protein (AMWAP), forward primer 5′-tttgatcactgtggggatga-3′, reverse primer 5′-acactttctggtgaaggcttg-3′, probe #1; translocator protein (TSPO), forward primer 5′-actgtattcagccatggggta-3′, reverse primer 5′-accatagcgtcctctgtgaaa-3′, probe #33; IL6, forward primer 5′-gatggatgctaccaaactggat-3′, reverse primer 5′-ccaggtagctatggtactccaga-3′, probe #6; inducible nitric oxide synthase (iNOS), forward primer 5′-ctttgccacggacgagac-3′, reverse primer 5′-tcattgtactctgagggctga-3′, probe #13; CCL2, forward primer 5′-catccacgtgttggctca-3′, reverse primer 5′-gatcatcttgctggtgaatgagt-3′, probe #62; CASP8, forward primer 5′-tgaacaatgagatccccaaat-3′, reverse primer 5′-caaaaatttcaagcaggctca-3′, probe #11; and ATP5B, forward primer 5′-ggcacaatgcaggaaagg-3′, reverse primer 5′-tcagcaggcacatagatagcc-3′, probe #77. Measurements were performed in triplicates. ATP5B expression was used as reference gene and results were analyzed with the ABI sequence detector software version 2.4 using the ΔΔCt method for relative quantification.
Library preparation and RNA-seq were carried out as described in the Illumina TruSeq Stranded mRNA Sample Preparation Guide, the Illumina HiSeq 1000 System User Guide (Illumina, Inc., San Diego, CA, USA), and the KAPA Library Quantification Kit—Illumina/ABI Prism User Guide (Kapa Biosystems, Inc., Woburn, MA, USA). In brief, 300 ng of total RNA was used for purifying the poly-A containing messenger RNA (mRNA) molecules using poly-T oligo-attached magnetic beads. Following purification, the mRNA was fragmented to an average insert size of 200–400 bases using divalent cations under elevated temperature (94 °C for 4 min). The cleaved RNA fragments were copied into first-strand cDNA using reverse transcriptase and random primers. Strand specificity was achieved by replacing dTTP with dUTP in the Second Strand Marking Mix (SMM), followed by second-strand cDNA synthesis using DNA Polymerase I and RNase H. The incorporation of dUTP in second-strand synthesis quenches the second strand during amplification, because the polymerase used in the assay is not incorporated past this nucleotide. The addition of Actinomycin D to First Stand Synthesis Act D mix (FSA) prevents spurious DNA-dependent synthesis, while allowing RNA-dependent synthesis, improving strand specificity. These cDNA fragments then had the addition of a single “A” base and subsequent ligation of the adapter. The products were purified and enriched with PCR to create the final cDNA library. The libraries were quantified using the KAPA SYBR FAST ABI Prism Library Quantification Kit. Equimolar amounts of each library were used for cluster generation on the cBot (TruSeq SR Cluster Kit v3). The sequencing run was performed on a HiSeq 1000 instrument using the indexed, 50 cycles single read (SR) protocol and the TruSeq SBS v3 Kit. Image analysis and base calling resulted in .bcl files, which were converted into .fastq files by the CASAVA1.8.2 software. Library preparation and RNA-seq were performed at the service facility “KFB—Center of Excellence for Fluorescent Bioanalytics” (Regensburg, Germany).
The RNA express workflow on Illumina BaseSpace was used to determine differential gene expression in two biological replicates of control retinas, light-exposed retinas, and light-exposed retinas treated with minocycline, respectively. In brief, alignment of RNA-seq reads to the mouse UCSC mm10 genome and mapping to genes was performed with STAR aligner [
35]. Differential gene expression between the two biological replicates each was calculated with DESeq2 [
36]. The cutoff of genes considered to be differentially expressed was a log2 fold change of ≥2 or ≤−2.
Integrative analysis of genome-wide expression activities was performed with the Gene Expression Dynamics Inspector (GEDI), a Matlab (Mathworks, Natick, MA) freeware program which uses self-organizing maps (SOMs) to translate high-dimensional data into a 2D mosaic [
37]. Each tile of the mosaic represents an individual SOM cluster and is color-coded to represent high or low expression of the cluster’s genes, thus identifying the underlying pattern.
The RNA-seq raw data and normalized DESeq2 counts of this study are publicly available at the National Center for Biotechnology Information Gene Expression Omnibus (
http://www.ncbi.nlm.nih.gov/geo/) as series record GSE71025.
Nitrite measurement
Nitric oxide concentrations were determined by measurement of nitrite released into BV-2 culture supernatants using the Griess reagent system (Promega). Fifty-microliter cell culture supernatants from differentially stimulated BV-2 cells were incubated with 100 μl Griess reagent in each well of a translucent 96-well plate. After incubation for 30 min at room temperature, absorbance was measured at 540 nm on an Infinite F200 Pro plate reader (Tecan). Nitrite concentrations were calculated on the basis of a sodium nitrite reference curve.
661 W photoreceptor apoptosis assay
To investigate microglial neurotoxicity and to assess whether minocycline also has direct effect on photoreceptors, 661W cells were incubated for 48 h with culture supernatants from differentially stimulated BV-2 cells or culture supernatants supplemented with minocycline. Apoptotic cell death was determined using the Caspase-Glo® 3/7 Assay (Promega). Cells were lysed and incubated with a luminogenic caspase-3/7 substrate, which contains the tetrapeptide sequence DEVD. After incubation at room temperature for 1 h, the generated luminescence was measured on an Infinite F200 Pro plate reader (Tecan). A blank reaction without cells was used to determine background luminescence associated with the cell culture system and Caspase-Glo® 3/7 reagent. The values of the blank reactions were subtracted from all experimental values. Negative control reactions were performed to determine the basal caspase activity of 661 W cells. Relative luciferase units (RLU) reflect the level of apoptotic cell death.
Animals
All experiments were performed with 10–14-weeks-old albino BALB/cJ mice of both sexes. Animals were housed in an air-conditioned environment with 12-h light-dark schedule and had free access to water and food. All experimental procedures complied with the German law on animal protection and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The animal protocols used in this study were reviewed and approved by the governmental body responsible for animal welfare in the state of Nordrhein-Westfalen (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen, Germany) (reference number 84-02.04.2015-A039).
Minocycline administration
The mice received intraperitoneal injections of minocycline at a dose of 45 mg/kg or NaCl as solvent control where indicated twice daily for the first two days, starting one day before the light exposure and once daily for the remaining days.
Light exposure regimen
BALB/cJ mice were dark-adapted for 16 h before light exposure. After pupil dilatation with 1 % Phenylephrin and 2.5 % Tropicamid under dim red light, the mice were exposed to bright white light with an intensity of 15.000 lx for 1 h. After light exposure, the animals were housed in dark-reared conditions overnight and then maintained under normal light conditions for the remaining experimental period.
Immunohistochemistry
Eyes were harvested for immunohistochemical analysis 4 days after light exposure. After fixation with 4 % paraformaldehyde, eyes were embedded in optimal cutting temperature (OCT) compound or dissected for retinal flat mount analysis. Sixteen-micrometer sections were rehydrated with PBS and blocked with dried milk solution. Flat mounts were incubated with 5 % Tween, 5 % Triton-X100 in PBS overnight and non-specific binding was blocked by incubation with dried milk solution. Subsequently, retinal sections and flat mounts were incubated with primary antibodies at 4 °C overnight. Primary antibodies targeting the following molecules were used: rabbit anti-Iba1 antibody (Wako Chemicals, Neuss, Germany), rabbit anti-TSPO antibody (Abcam, Cambridge, UK), and rabbit anti-Glial Fibrillary Acidic Protein (G9269, Sigma). After a washing step, the sections and flat mounts were incubated with a secondary antibody either conjugated to Alexa488 (green) or Alexa594 (red) (Jackson Immuno-Research, West Grove, PA, USA) for 1 h. After counterstaining with 4',6-diamidino-2-phenylindole (DAPI), the samples were mounted in DAKO fluorescent mounting medium (Dako Deutschland GmbH, Hamburg, Germany) and analyzed with an Axioskop2 MOT Plus Apotome microscope (Carl Zeiss).
Optical coherence tomography (OCT)
Animals were anesthetized by intraperitoneal injection of Rompun (10 mg/kg body weight)-Ketavet (100 mg/kg body weight), and their pupils were dilated with Phenylephrine HCl (0.25 %)-Tropicamide (0.05 %) before image acquisition. Spectral-domain OCT (SD-OCT) was performed on both eyes with a Spectralis™ HRA + OCT device (Heidelberg Engineering) to investigate structural changes in the retina after light exposure and minocycline administration. Thickness measurements were performed with a circular ring scan (circle diameter 1, 3, 6 ETDRS), centered on the optic nerve head, which represents the average retinal thickness [μm] in a certain field. Central retinal thickness was calculated from four fields around the optic nerve head using the Heidelberg Eye Explorer Software.
Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay
Retinal sections were labeled with an in situ cell death detection kit, Fluorescein (Roche) to detect the amount of apoptotic cells 4 days after light exposure. For a better overview, the sections were also counterstained with DAPI for 10 min. After a washing procedure, sections were mounted in DAKO fluorescent mounting medium (Dako Deutschland GmbH, Hamburg, Germany) and analyzed with an Axioskop2 MOT Plus Apotome microscope (Carl Zeiss).
Statistical analysis
Nitrite secretion, caspase 3/7 activities, and real-time RT-PCR data from BV-2 cells were analyzed with one-way ANOVA and Bonferroni’s multiple comparison tests. Real-time RT-PCR data of murine retinas and OCT measurements of retinal thickness were analyzed using a one-way ANOVA and Dunnett’s multiple comparison tests. p < 0.05 was considered statistically significant.
Discussion
Several studies revealed that minocycline exerts anti-microbial, anti-inflammatory, anti-apoptotic, and neuroprotective properties in different animal models of neuronal degenerative diseases, including Parkinson’s disease, multiple sclerosis, Alzheimer’s disease, Huntington’s disease, amyotrophic lateral sclerosis, and retinitis pigmentosa [
16,
18‐
20,
23,
25,
42‐
44]. Here, we report for the first time that minocycline administration preserves the retinal structure in a white light-induced degeneration model, which mimics several features of dry AMD and monogenic retinal dystrophies. The main questions discussed in this study were if and how minocycline stimulation modulates the neurotoxic potential of microglia and whether minocycline administration could be a suitable immuno-modulatory strategy to treat retinal degeneration.
Our in vitro studies in LPS-activated BV-2 microglia demonstrated that minocycline was capable to diminish gene expression of the pro-inflammatory markers CCL2, IL6, and iNOS in a rapid temporal response. This is consistent with earlier reports with microglia isolated from rat retinas where minocycline dampened mRNA expression and secretion of IL1β, TNFα, and iNOS [
45]. In line with this, Henry et al. reported that minocycline reduced mRNA levels of TLR 2, MHC-II, IL1β, and IL6 in BV-2 cells [
46]. These results together with our data suggest that minocycline is highly effective in reducing pro-inflammatory gene transcription in microglia.
The production of toxic nitric oxide by phagocytes plays an important role in degenerative processes [
47‐
49]. Here, we demonstrated that minocycline effectively dampened nitric oxide production of LPS-stimulated microglia. Furthermore, minocycline reduced the general neurotoxic potential of reactive microglia on 661W photoreceptor cells. Supernatants from microglia that were stimulated with both LPS and minocycline were less toxic to photoreceptors than those of cells that were treated with LPS only. Investigations in RAW264.7 macrophages and spinal cord microglia also confirmed that minocycline blocks the production of toxic nitric oxide [
50,
51]. Hence, others have found an association of reduced iNOS levels and NO itself with delayed neurodegeneration [
19,
47,
52]. In the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) mouse model of Parkinson’s disease, minocycline reduced MPP+ induced glial iNOS expression, which was most likely due to an inhibition of p38 MAPK phosphorylation [
19].
Besides the reduction of NO production, several other mechanisms of action are discussed for the beneficial effects of minocycline. Indeed, we here also showed that minocycline administration directly protected photoreceptors from microglia mediated death. In a model of Huntington’s disease, minocycline delayed disease progression and mortality by inhibition of caspase-1 and caspase-3 up-regulation, enzymes which play an important role in the induction of necrosis and apoptosis [
43]. Studies in R6/2 mice confirmed that minocycline can inhibit both caspase-independent (apoptosis inducing factor) and caspase-dependent (Smac/Diablo and cytochrome c) mitochondrial cell death pathways [
53]. However, despite its ability to reduce pro-inflammatory markers in a mouse model for age-related neuron loss, minocycline failed to inhibit apoptosis and neuron loss [
54].
A further mechanism how minocycline exerts its protective functions is the inhibition of matrix metalloproteases (MMPs). MMPs are a family of zinc- and calcium-dependent proteolytic enzymes that are responsible for the degradation of structural proteins in the extracellular matrix and their activation is associated with neurological disorders. Thus, blockage of MMPs by minocycline decreased the cerebral infarct size in a mouse model of focal cerebral ischemia. This effect was most likely due to the inhibition of CCL2, TNFα, and indoleamine 2,3-dioxygenase (IDO) expression [
55,
56].
Our OCT images and immunohistochemical analysis showed that the treatment with minocycline preserved retinal structure and reduced the amount of apoptotic cells after exposure to bright white light. The protection of photoreceptors was associated with a reduced number of reactive, amoeboid-shaped, TSPO-positive microglia in the outer retina. Accordingly, our mRNA analyses of retinas revealed suppression of microglial activation markers including AMWAP and CD68 after administration of minocycline [
38‐
40]. Analysis at different time points after light exposure revealed that microglial activation is an early event that precedes photoreceptor death. Moreover, our RNA-sequencing data revealed a global transcriptomic effect of minocycline with a complete cluster of light-damage-induced pro-inflammatory genes that was counter-regulated by minocycline. Of note, the RNA-seq dataset also detected up-regulation of structural retinal genes such as keratins and adhesion molecules by minocycline. These results strongly indicate that modulation of microglia reactivity is a key mechanism in minocycline’s mode of action in the retina. This notion is supported by studies demonstrating that inhibition of microglia with minocycline protects from neuronal degenerative diseases including Alzheimer disease or Parkinson disease [
16,
18]. In the eye, minocycline could also reduce microglial activation and improve neuronal function [
13,
25,
57,
58]. In a model of green light exposure, preservation of retinal structure and ERG amplitudes by minocycline was associated with reduced numbers of CD11b+ cells in the outer retina [
58]. To mimic exposure to bright daylight, which is discussed as a contributing factor for retinal degenerations, we used 15.000 lux UV-free white light in our studies. White light has an emission spectrum similar to that of daylight, and it is less artificial than light of a particular wavelength [
59,
60]. Furthermore, white light contains also short wavelength blue light (403 nm) which is thought to have a higher damaging potential than light of longer wavelength including green light (490–580 nm), which was used in earlier studies [
58]. The higher damaging potential of blue light is due to a process called photo-reversal, the regeneration of rhodopsin from bleaching intermediates that results in a higher number of photon absorption in a certain time span. Because of different light exposure settings in our study and the work of Zhang et al. [
58], different underlying damage mechanisms cannot be ruled out [
28,
59,
61,
62].
In the
rd10 mouse model for retinitis pigmentosa, minocycline inhibited microglial activation and down-regulated the expression of pro-inflammatory molecules including TNFα, COX1, and COX2. Moreover, pro-apoptotic molecules such as BAX and Caspase 3 were suppressed by minocycline, and the retinal structure and function were preserved [
25]. Of note, minocycline could also diminish photoreceptor death in
rds mice by a microglia-independent mechanism as depletion of microglia by clodronate prevented their recruitment but failed to inhibit photoreceptor apoptosis [
63].
Competing interests
All authors declare no competing financial interests.
Authors’ contributions
RS designed the research, performed the research, analyzed the data, and wrote the manuscript. MS, AC, TS, and CM performed the research and analyzed the data. TL analyzed the data, obtained the funding, designed the study, and wrote the manuscript. All authors read and approved the final manuscript.