Effects of aging and environmental tobacco smoke exposure on ocular and plasma circulatory microRNAs in the Rhesus macaque

Animals

Rhesus macaques used for this study were housed at the California National Primate Research Center (CNPRC) at University of California, Davis. There were four experimental groups of female animals (a total of 16 animals): Group 1 (young, 1–3 years; n=4), Group 2 (old, 19–28 years; n=5), Group 3 (middle aged, 10–16; n=3), and Group 4 (middle aged, 9–14; n=4). The Group 4 was exposed to smoke for 1 month at the CNPRC facility (Appendix 1). All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. These procedures were also authorized by the Institutional Animal Care and Use Committee (#18988) at the University of California, Davis.

Clinical assessment of animals

Rhesus macaques selected for the ETS exposure experiment were clinically assessed and found healthy in the absence of (n=6) or following (n=4) ETS exposure. Fundus photography using the CF-1 retinal camera (Canon, Tokyo, Japan) and spectral domain optical coherence tomography (SD-OCT) imaging were performed using the Spectralis device (Heidelberg Engineering, Heidelberg, Germany) as detailed below.

Spectral domain OCT Imaging

SD-OCT imaging was performed using the Spectralis SD-OCT device (Heidelberg Engineering), which had been modified with a flat chin-rest to allow the Rhesus monkeys’ heads to be positioned. OCT images were captured with a 30° × 5° SD-OCT raster scan with 1536 A-scans per B-scan and 234 μm of spacing between B-scans. This was performed in the high-resolution mode with 25 scans averaged for each B-scan using the device’s eye-tracking capability. Images were semiautomatically segmented using the Duke Optical Coherence Tomography Retinal Analysis Program (DOCTRAP, version 62.0), custom image analysis software designed using MATLAB (Mathworks) [3]. Segmentation boundaries were automatically calculated by DOCTRAP using graph-based determination of reflectivity profiles and then manually refined by the two graders. Thickness measurements were averaged across the 3 mm segment centered on the fovea in each horizontal line-scan, and included the nerve fiber layer (NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), photoreceptor inner segments (IS), photoreceptor outer segments (OS), retinal pigment epithelium (RPE), choriocapillaris (CC), and choroid (C). Total retinal thickness was measured from the internal limiting membrane (ILM) to Bruch’s membrane (BM).

Statistical analysis

All measurements reported in this study were calculated using the average measurements from two graders. The mixed-effect ANOVA was performed to test the pre/post exposure effects on all retinal layers. A mixed-design ANOVA model (also known as a split-plot ANOVA) is used to test for differences between two or more independent groups while subjecting participants to repeated measures. Differences in chorioretinal layer thicknesses obtained from the SD-OCT pre/post exposure were compared using a mixed-effect ANOVA, with two independent variables: The within-subject factor variable was pre/post exposure and eyes [oculus sinister (OS) / oculus dextrus (OD)] while the between-subject factor variable was an animal ID. The dependent variable was the thickness of each layer. All statistical analysis was performed using SPSS software (version 22, IBM). The segmentation analysis and quantification of the retinal layers are shown for the total retinal thickness measurement along the central 3 mm region for all four animals. The mean value of retinal layer thickness was calculated for each layer in four monkeys pre/post exposure. No statistical significance was found for pre/post exposure eyes.

Tobacco smoke exposure study design

Groups 1 through 3 were housed in standard indoor animal housing, while Group 4 was housed in standard caging in a modified animal holding room specifically designed for ETS. Tobacco smoke exposure occurred over 4 weeks with ETS delivered for approximately 6 h per day, 5 days a week. The exposure room was ventilated with filtered air during non-ETS exposure periods or with ETS diluted with filtered air during exposure periods. Necropsies were performed on the day immediately following the last exposure day.

Tobacco smoke exposure study analyses

Tissue collection

Ocular fluid samples were collected when the animals were euthanized. Blood samples were collected at the beginning of the experiment and when the animals were euthanized. Animals were euthanized following the standard CNPRC protocol: the animal was sedated with Ketamine at a dosage of 10 mg/kg. After that sodium pentobarbital greater or equal to 120 mg/kg Fatal Plus or Beuthanasia Plus 1 ml/4.5 kg was administered intravenously to the animal for euthanasia. After each animal was euthanized, 100–200 μl samples of aqueous and vitreous humor and 10 ml of venous blood for plasma were collected. Plasma was isolated using ACCUSPIN tubes (MilliporeSigma, St. Louis, MO) following the manufacturer’s protocol. Plasma samples were stored at −80 °C.

Ocular tissue fixation and hematoxylin-eosin staining

Enucleated eyes were immersion fixed overnight at room temperature in 2% phosphate-buffered paraformaldehyde containing 0.5% glutaraldehyde, pH 7.4 (0.1 M phosphate buffer). After overnight fixation, the eyes were bisected, the lens removed, and the bisected eye was placed in 4% phosphate-buffered formalin for approximately 1 week to reverse glutaraldehyde fixation before embedding in paraffin. The eyes were embedded in paraffin as follows: washed in PBS (1X; 154 mM NaCl, 1.1 mM KH₂PO₄, 5.6 mM Na₂HP0₄, pH 7.4), dehydrated through graded ethanol washes (50%, 75%, 95%, and 100% ethanol), and gradually cleared in xylene using a step-wise procedure (2:1 ethanol to xylene; 1:1 ethanol to xylene; 1:2 ethanol to xylene), followed by xylene before embedding in paraffin. Sections were cut (6 μm) using a Leica RM2125RT microtome (Leica, Nussloch, Germany), placed on SuperFrost Plus microscope slides, and dried overnight at room temperature. Standard hematoxylin-eosin staining was performed on the sectioned eyes.

RNA isolation

All the samples were collected within a 1 h after the animals were euthanized, processed immediately, and stored in aliquots at −80 °C until RNA isolation. The total RNA, which included miRNAs, was isolated from each of the individual samples using miRCURY™ RNA Isolation Kits -Biofluids (Qiagen, Germantown, MD). A small aliquot of the sample was frozen separately for quantification and quality control on BioAnalyzer, so the main sample tube was thawed only once during the process of labeling probe for GeneChip microarray hybridization. Total isolated RNA (5–10 ng/μl) from fluid samples was analyzed using the BioAnalyzer using Small RNA Analysis Kit (Agilent Technologies, Santa Clara, CA) following the manufacturer’s protocol.

Probe labeling for hybridization to microRNA microarrays

MiRNA samples (20 ng) were labeled as probes, using the FlashTag kit (ThermoFisher Scientific, Waltham, MA) following the manufacturer’s protocol and hybridized to microRNA microarrays. Two Affymetrix miRNA microarrays GeneTitan Plates 4.0, each with 24 samples, for a total of 48 samples, were run at the UCSF Institute for Human Genetics (IHG). Raw data in the form of probe cell intensity data (.cel) files were retrieved and subjected to bioinformatics analysis.

Bioinformatics analysis

Probe cell intensity data (.cel) from Affymetrix GeneChip® miRNA Arrays were analyzed with the Affymetrix® Expression Console (EC) software (ThermoFisher Scientific). The application uses the RMA + DABG analysis for doing background subtraction, normalization, and summarization of the probe sets from Affymetrix expression microarrays, and the final output is a .chp file. Spearman rank order correlation coefficients between biologic replicates were computed by EC to assess the between samples correlations (Appendix 2). Generated .chp files for each sample were imported into Affymetrix Transcriptome Analysis Console (TAC; ThermoFisher Scientific) for subsequent analysis of differentially expressed genes.

Quantitative polymerase chain reaction (qPCR) assays

Real-time qPCR assays (TaqMan assays, ThermoFisher Scientific) were run on miRNA isolated from plasma samples for ETS versus control on three replicas for each group and in triplicate for each replica. Assays were run at the Real-time PCR Research and Diagnostics Core Facility at UC Davis. Each miRNA sample was amplified using a TaqMan preamplification procedure and TaqMan preamp primers (Applied Biosystems, Life Technologies, Grand Island, NY) following manufacturers protocols. Reverse transcription thermal-cycling conditions were: 16 °C/30 min, 42 °C /30 min, 85 °C /5 min, hold 4 °C /∞. Pre-amplification thermal-cycling conditions were: 95 °C /10 min, 55 °C /2 min, 72 °C /2 min, 95 °C /15 s, 12 cycles of 95 °C /15 s, 60 °C /4 min, hold at 99.9 °C /10 min, hold 4 °C /∞. The samples were run in 384 well plates. Real-time thermal-cycling conditions: hold 95 °C /10 min, 40 cycles of 95 °C /15 sec, 60 °C /60 s, hold 4°C /∞. The TaqMan assays performed were miR-221 (ABI No. 000524, hsa-miR-221), miR-26a (ABI No. 000405, hsa-miR-26a), let-7c-5p (ABI No. 000379, hsa-let-7c), and three control genes, miR-486-5p (ABI No. 001278, hsa-miR-486), miR-320b (ABI No. 002844, hsa-miR-320b), and miR-638 (ABI No. 001582, hsa-miR-638). MiRNA assays were conducted at a dilution of 1:100s, and each run was done in triplicate. A control with no enzyme in the reverse transcriptase (RT) step was included as a negative control.

Tobacco smoke exposure

The adult female Rhesus macaques in Group 4 (exposure) received four 7-day cycles of 1.42–2.15 mg/m³ of tobacco smoke for an average of 6 h/day on days 1–5, followed by 2 days in the filtered air (Table 1). During exposure, filtered air ventilation was lowered to 3.5 room air exchanges per hour. The daily average nicotine level of exposure measured through XAD-4 sorbent tubes was (0.610 mg/m³ ±0.090). In animals that underwent ETS exposure, ocular health was evaluated by veterinary and physician ophthalmologists (SMT and GY) pre- and post-exposure. The control group (Group 3) was housed in standard indoor animal housing.

Cotinine results

Cotinine is an alkaloid found in tobacco and is the predominant metabolite of nicotine [5]. Cotinine concentrations in the control and ETS-exposed animals at day 30 are displayed in Table 2. On average, ETS-exposed animals contained statistically significantly higher (p<0.01), concentrations of cotinine compared to control animals at the 15-day time point (48.6±13.1 ng/ml versus 0 ng/ml) and the 30-day time point (54.9±5.0 ng/ml versus 0.06±.001 ng/ml).

Clinical evaluations of ETS-exposed macaques

Rhesus macaques from Groups 3 and 4 were clinically assessed before and following ETS exposure. Ophthalmic examinations did not identify any visible ocular changes pre- and post-exposure (Figure 1). Retinal imaging, performed using OCT, followed by segmentation analysis and quantification of the retinal layers is shown for the total retinal thickness measurement along the central 3 mm region for all four animals (Table 3). There were trends noted in the thinning of the nuclear fiber layer (NFL; 17.36±0.860 um versus 16.01±0.470 um (p<0.17) and in the thickening of the choriocapillaris layer (CC; 18.60±2.050 versus 19.46±2.240 (p<0.34), pre- and post-ETS exposure, although these trends were not statistically significant.

Small RNA profiles from plasma and aqueous and vitreous humor

Representative profiles from the BioAnalyzer small RNA kit are shown in Figure 2. There was an abundance of circulatory miRNA molecules in each of the fluids analyzed. Representation of miRNA in each of the fluid compartments isolated from 200 μl of fluid was within the range of 2.30–14.6 ng/μl in the aqueous, 0.8–2.2 ng/μl in the vitreous, and 0.16–0.32 ng/μl in the plasma. The small RNA bands were abundant (Figure 2, red arrow) and the range of the sizes of the small RNA molecules was comparable to the range in the profile of the high-quality small RNA fraction in the solid tissue RNA prep (data not shown). The total concentration of miRNA in nanograms/microliters did not change statistically significantly in the plasma samples from the same animals pre- and post-exposure.

The total number of miRNAs detected

In the plasma samples, 40 microRNAs were ubiquitously present (p<0.05) in all samples, and up to 204 variably. In vitreous, ten miRNAs were ubiquitously present in all samples and up to 88 variably, and in aqueous, seven were ubiquitously present and up to 148 variably. The presence of microRNAs was established according to the Affymetrix P/A/M detection algorithm based on the Wilcox signed-rank test introduced by Affymetrix [6].

Changes in primate circulatory miRNA expression after ETS exposure

In the plasma samples, 45 miRNAs were statistically significantly upregulated after ETS exposure (Figure 3A). In the vitreous samples, three miRNAs were statistically significantly downregulated after ETS exposure (Figure 3B). There were no statistically significantly dysregulated miRNAs observed in the aqueous samples. All miRNAs with their fold change and p-values (−1.5 ≥ FC ≥ 1.5; p<0.05) are listed in Appendix 3, Appendix 4, Appendix 5, Appendix 6, Appendix 7.

Age-related changes in circulatory miRNAs

In the plasma samples from the aging monkeys, two miRNAs were statistically significantly downregulated between old (Group 2) versus young (Group 1) macaques (Figure 4). In the vitreous and aqueous samples, six miRNAs were statistically significantly downregulated in each compartment (−1.5 ≥ FC ≥ 1.5; p<0.05). All dysregulated miRNAs differed between the vitreous and aqueous samples.

Validation of microarray results with qPCR

Independent quantitative PCR (qPCR) validation of representative differentially expressed genes was performed using commercial gene expression assays (TaqMan; Applied Biosystems), following manufacturers protocol (for detailed reaction conditions see Materials and Methods section on Quantitative polymerase chain reaction (qPCR) assays). The fold changes and their direction were confirmed for the chosen sets of genes. Data are presented in Table 4.

Pathway analysis of miRNAs differentially regulated after ETS exposure

Ingenuity Pathway Analysis identified two top networks that are activated during ETS exposure: IL-17A and VEGF. One of the top IPA pathways that incorporates the activation of both networks is the inflammatory pathway of recruitment and activation of eosinophils, Th2 lymphocytes, and macrophages (Figure 5).

Primate ETS exposure levels were within the range of strong environmental tobacco smoke exposure

Curtis’s and other groups have shown that 1 month of high ETS exposure produces chronic obstructive pulmonary disease (COPD) in primate animal models [12,13]. Our working hypothesis was that the retina and the choroid, highly vascular tissues that are sensitive to oxidative stress, might be similarly impacted by short-term ETS exposure. Furthermore, environmental second-hand smoke has been shown to have serious health effects [14-16]. We measured animal exposure to ETS using several criteria: particulate matter, nicotine, and cotinine. Cotinine is an alkaloid found in tobacco and the predominant metabolite of nicotine [5]. Cotinine has an in vivo half-life of approximately 20 h and is typically detectable up to 1 week after the use of tobacco. The amounts of cotinine in blood, saliva, and urine are proportionate to the amount of exposure to tobacco smoke; thus, cotinine is a valuable indicator of tobacco smoke exposure, including secondary (passive) smoke [17]. The nicotine concentrations measured in the current study were consistent with second-hand smoke exposure, and the cotinine assays confirmed marked ETS exposure with a concentration of about 50 ng/ml. In humans, cotinine concentrations of about 30 ng/ml indicate marked exposure [18].

The thickness of retinal layers did not change statistically significantly following ETS exposure, which was consistent with the literature on humans

All macaques were clinically assessed before and following ETS exposure. Fundus photography of the back of the eye did not show any visible changes pre- or post-exposure. Retinal imaging using OCT, followed by segmentation analysis and quantification of the retinal layers, showed no statistical significance in the total retinal thickness or any of the retinal layers. This result was consistent with findings in human smokers, where the central macular thickness shows no statistically significant differences between smokers and non-smokers [19,20]. Although not reaching a statistical significance level, perhaps because of the small animal numbers and only 1 month of exposure, a trend in the thinning of the NFL was observed. This observation is consistent with the literature showing that smoking tobacco may cause considerable thinning of the NFL in heavy smokers [20-22]. In addition, we observed the trend of choriocapillaris thickening in primates exposed to ETS. This result is consistent with the findings that exposure to ETS results in the formation of sub-RPE deposits, thickening of Bruch’s membrane, and the accumulation of deposits within Bruch’s membrane. All these processes lead to an injury stimulus of the choriocapillaris and RPE, which may explain the association between tobacco smoking and early AMD [23]. One of the limitations of this study is that the OCT-based measurements are limited by the inclusion of Henle’s fiber layer within the ONL rather than the OPL to which the layer belongs [24]. Thus, true longitudinal changes in the OPL at the macula may not be accurate.

Circulatory miRNA changes in response to ETS exposure were different from those that occur during the normal aging process

Upon ETS exposure, the greatest changes in circulatory miRNAs were observed in the plasma (45 miRNAs were strongly upregulated), followed by the vitreous (three miRNAs were downregulated). However, there were no statistically significantly dysregulated miRNAs detected in the aqueous humor. These results differ markedly from what occurs during normal aging where only two miRNAs were statistically significantly downregulated in the plasma, while in the vitreous and aqueous six different miRNAs were statistically significantly downregulated (-1.5>FC>1.5; p<0.05). The only exceptions, two out of three novel miRNAs identified as downregulated following ETS exposure in the vitreous (miR-6794 and miR-6790) were the same ones observed to be downregulated in the vitreous during normal aging.

Dysregulated miRNAs common to ETS exposure and AMD

Circulating miRNAs are excellent biomarkers. They can be easily sampled from plasma and bodily fluids, including ocular fluids, thus gaining the name of the liquid biomarker or liquid biopsy [25]. They are stable for hours, even at room temperature, because they travel protected associated with proteins or encapsulated by small vesicles called exosomes [26].

In addition to being accessible biomarkers, miRNAs are biologically powerful molecules that regulate post-transcriptional gene expression. MiRNA expression is necessary for cellular function, such as cell proliferation, apoptosis, inflammation, and a wide variety of cellular processes. MiRNAs function through base-pairing with mRNA, which leads to degradation or impaired translation of that particular mRNA [27]. Therefore, they can be referred to as “dimmer switches” for gene expression, because of their ability to repress gene expression without completely silencing it. Six of the miRNAs identified in the present study in primates exposed to ETS (let-7a, let-7c, miR-106a-5p, miR-106b-5p, miR-17-5p, and miR-361-5p) were also identified as circulatory candidate miRNAs in AMD [28-31]. The let-7 family and miR-361-5p are upregulated in retinal tissue in advanced AMD, targeting TGFBR1 [29,32]. MiR-17-5p has been shown to be upregulated in the plasma of patients with wet AMD [33]. MiR-17-5p negatively regulates TGFBR2 expression by directly binding to the 3′ untranslated region (UTR) of TGFBR2 mRNA, thus promoting cell growth and migration in cancer [34]. MiR-106 has been identified in a cohort of miRNAs that respond to oxidative stress exposure in ARPE-19 cells [35]. MiR-106a has been shown to target PTEN mRNA and lncRNA LINC00657, thus playing an important role in the regulation of PTEN expression in cancer [36].

Pathway analysis of dysregulated miRNAs after ETS exposure of Rhesus macaques targeted two signaling networks IL-17 and VEGF, and the inflammatory macrophage activation and recruitment pathway

Ingenuity Pathway Analysis of candidate miRNAs and their target genes identified two top networks that were activated during ETS exposure: IL-17A and VEGF. Upregulation of IL-17A has been previously observed with ETS exposure. Specifically, Siew and colleagues demonstrated that ETS increases bronchial mucosal IL-17A expression, and neutrophilic inflammation in asthmatics [37]. Levanen and colleagues implied that long-term exposure to tobacco smoke increases extracellular cytokine signaling via IL-17A in the peripheral airways of smokers with COPD and that this signaling may involve cytotoxic T-cells [38]. Interestingly, IL-17A also appears to be important in the pathogenesis of AMD. For example, genetic variants of IL-17A have been functionally associated with an increased risk of AMD [39]. Zhang and colleagues suggested that IL-17A triggers a key inflammatory mediator, IL-1β, in RPE cells, via NLRP3 inflammasome activation [40]. Therefore, the IL-17A pathway may be an important therapeutic target for AMD. Another study demonstrated that a functional single nucleotide polymorphism in the IL-17A 3′ UTR is targeted by miRNA in vitro and that miRNA-mediated gene dysregulation may play a role in the pathogenesis of AMD [41].

ETS exposure causes extensive tissue remodeling in the upper airways, and VEGF plays a critical role in this process. It has been shown that ETS extract has a stimulatory effect on VEGF expression through the TLR4/ROS/MAPKs/NF-kappaB signaling pathway in nasal fibroblasts [42]. In addition, a human RPE cell line (ARPE-19) exposed to an ETS concentrate resulted in premature senescence, thus supporting the role of oxidative damage in ETS-induced senescence activation. Furthermore, the senescent ARPE-19 cells upregulated VEGF and simultaneously downregulated complement factor H expression. As these phenomena are involved in AMD pathogenesis, these studies coupled with the present results support the hypothesis that ETS-induced senescence activation is important in the induction and progression of AMD. Moreover, this hypothesis would also explain the striking association of AMD pathogenesis with tobacco smoking [43].

One of the top IPA pathways that incorporates activation of both networks is the inflammatory recruitment and activation of eosinophils, Th2 lymphocytes, and macrophages pathway. This result fits with studies performed on infant monkeys after perinatal ETS exposure, which showed a statistically significant increase in mast cell, eosinophil, monocyte, and lymphocyte cell numbers in the lungs of infant monkeys [44]. It has been demonstrated that ETS exposure during early life enhances local Th2 immunity by impairing normal pulmonary Th1 immune maturation. This effect was greater in animals that began ETS exposure in utero compared to those that were exposed postnatal [45]. However, an association of the inflammatory response and two distinct types of immune responses—innate (complement system, phagocytes, mast cells, eosinophils, macrophages) and adaptive (T- and B-lymphocytes)—have been strongly implicated in AMD. An overactive complement system is associated with AMD pathogenesis, and serum proinflammatory cytokines, including IL-17, are elevated in patients with AMD. IL-17 is produced by complement C5a-receptor-expressing T-cells, such as gamma-delta T-cells, members of the innate immune response [46-48]. VEGF plays an important role in pathological angiogenesis, and anti-VEGF agents have been used to treat ocular diseases that are driven by pathological angiogenesis [49-51]. Based on the present results from the ETS-exposed monkeys and literature published on AMD, it appears likely that activation of IL-17A and VEGF networks, as well as inflammatory pathway of recruitment and activation of eosinophils, Th2 lymphocytes, and macrophages, are common for tobacco smoke and AMD pathogenesis.

In conclusion, this study identified changes in circulatory miRNAs in the plasma and ocular fluids of the Rhesus macaque following ETS exposure, which differed markedly in comparison to age-related changes. The results of this study suggest that ETS exposure may be an appropriate animal model for early dry AMD. The identification and characterization of eye-specific miRNA in ocular tissue, as well as their target molecular pathways, will generate opportunities to identify novel strategies for therapeutic and environmental modulation or intervention for blinding ocular diseases.