Differential proteomes in the vitreous of mice with a chronic hypoxic response in photoreceptors
Reduced oxygenation of photoreceptors and the RPE in the ageing eye may be a risk factor for the development of both neovascular and dry AMD. Chronic activation of the molecular response to hypoxia in RPE or photoreceptors leads to retinal degeneration that depends on hypoxia inducible transcription factors. For approaching the identification of accessible markers characteristic for photoreceptors with an activated hypoxic response, we used a proteomics approach to determine the protein composition of the vitreous humor in mice. To discriminate between rod and cone-specific effects we used genetically modified mice that had the hypoxic response activated specifically in rods of a rod-dominant retina (rodΔVhl) or a genetically engineered all-cone retina (coneΔVhl). For comparison, we used wild-type mice exposed for 6 hours to acute hypoxia. We identified 1,357 unique proteins in the vitreous of mice after acute hypoxia, 1,624 in rodΔVhl and 1,895 in coneΔVhl mice. Of these, 1,043 proteins were common to all three types of mice. Of the identified proteins, 257 were significantly regulated by a factor of 1.5 or more in hypoxic mice, 258 in rodΔVhl and 356 in coneΔVhl mice in at least one of the three analyzed time points. Only 51 of the significantly regulated proteins were common to the vitreous of rodΔVhl and coneΔVhl mice, suggesting different consequences of the activated hypoxic response for rods and cones. Guanylate binding protein 2 (GBP2) was found at increased levels in the vitreous of both rodΔVhl and coneΔVhl mice at all time points tested. This was also reflected by increased gene expression in the retina. Although retinal expression of the AMD-associated gene alpha-2 macroglobulin (A2M) appeared increased in both types of mice, the protein was only found elevated in the vitreous of rodΔVhl mice. Other proteins found increased included Serpina3n, synaptosome associated protein 25 (SNAP25) and others. The distinct protein compositions present at early and late time points, suggest a well-regulated process in our models. We hypothesize that some of the proteins identified at early time points may potentially be used as markers for the chronic hypoxic response of photoreceptors.
Sample Processing Protocol
Vitreous isolation The vitreous humor (VH) of mouse eyes was isolated as described previously (Skeie et al. 2011) with minor modifications. Briefly, animals were deeply anesthetized with Ketamin (130 mg/kg; Parke-Davis; Berlin; Germany) / Rompun (Xylazine; 26 mg/kg; Bayer AG; Leverkusen; Germany). After respiration has seized the heart was exposed and the animal perfused with 10 mL of Dulbecco’s PBS (DPBS; Thermo Fisher; Waltham; MA; U.S.A.) to minimize the risk of blood contamination in the VH during vitrectomy. To retrieve the VH, we removed the lens together with VH and retina through a slit in the cornea and transferred them to 50 µL of DPBS. The lens was then carefully separated from the retina causing the release of the VH into the solution. VH was separated from lens and retina by filtrating the solution through 0.1 µm centrifugal filters (Millipore Ultrafree; Merck & Cie; Schaffhausen; Switzerland) for 10 min at 5000 × g. The VH in the flow-through was directly snap frozen in liquid nitrogen and stored at – 80 °C until proteomic measurement. The retinal tissue retained by the filter was carefully separated from the lens and stored at – 80 °C. Sample preparation for proteomics VH samples were subjected to a filter assisted sample preparation (FASP)-digest adapted from Wiśniewski et al. (Wiśniewski et al. 2009). In brief, VH samples were denatured by 4 % SDS, 0.1 M dithiothreitol (DTT) and incubation at 95 °C for 5 min with subsequent sonication to disrupt collagen networks. Protein concentrations were determined by a Qubit fluorometric protein assay (Life Technologies; Zug; Switzerland). 20 µg of VH proteins were loaded onto a 30 kDa NMWL (Nominal Molecular Weight Limit) ultrafiltration centrifugal device (Microcon-30 kDa; Merck Millipore), centrifuged, washed with 8 M urea and alkylated with 0.05 M iodoacetamide (IAA) for 1 min. After membranes were washed 3 times with 8 M urea and twice with 0.5 M NaCl, proteins were digested over night at room temperature in 120 μL of 0.5 M triethylammonium bicarbonate buffer (pH 8.5) using trypsin (Promega; Dübendorf; Switzerland) at an enzyme to protein ratio of 1:100 (w/w). Digested samples were centrifuged, and eluted peptides acidified with 0.5 % trifluoroacetic acid (TFA) for the subsequent desalting step by C18 solid phase extraction columns (Sep-Pak Fenisterre; Waters Corp.; Milford; MA; U.S.A.) using 3 % acetonitrile (ACN) / 0.1 % TFA. Desalted tryptic peptides were lyophilized and resolubilized in 0.1 % formic acid (FA). Shot-gun proteomic analysis Analysis of samples using shot-gun proteomics was performed by liquid chromatography coupled tandem mass spectrometry (LC-MS/MS) on a high-resolution Fourier transformation mass spectrometer (QExactive; Thermo Fisher Scientific; Bremen; Germany). The mass spectrometer was interfaced to a nano-HPLC system (EASY-nLC 1000; Thermo Fisher Scientific) with a self-packed reverse-phase column (75 µm × 10 cm) containing C18 beads (AQ; 3 µm; 200 Å; Bischoff GmbH; Leonberg; Germany) at a flow rate of 200 nL × min-1. The column was equilibrated with 3 % ACN, 0.2 % FA in water. Peptides were eluted using the following gradient: 0 - 7 min, 3 % ACN, 7 - 127 min, 3 – 35 % ACN; 127 - 128 min, 35 – 95 % ACN; 128 - 130 min, 95 % ACN at a flow rate of 200 nL × min-1. The electrospray source was fitted with a 10 µm emitter tip (New Objective; Woburn; MA; USA) with an applied electric potential of 2.2 kV. High accuracy mass spectra were acquired in the mass range of 300 – 1,700 m/z and a target value of 3 × 106 ions in Orbitrap MS1, followed by higher energy collisional dissociation (HCD) fragmentation and top-12 MS2. Target ions already selected for MS2 were dynamically excluded for 30 s. General MS conditions were: normalized collision energy, 25 %; ion selection threshold, 5 × 104 counts; and activation time 120 ms for MS/MS acquisitions.
Data Processing Protocol
Protein identification and quantification For MS1 intensity-based label-free relative quantification, the ProgenesisQI for proteomics software (Vers. 3.0.5995, Nonlinear Dynamics Ltd., Tyne, UK) was used with high resolution ‘thermo.raw’ files and peptide charge states of 2+ to 5+ as input. The protein identification was based on MS2 peak lists generated by Mascot Distiller software (Vers. 2.5.1, Matrix Science Ltd., London, UK) with the following search settings: maximum missed cleavages: 2; peptide mass tolerance: 10 ppm; number of 13C = 1; and fragment ion tolerance: 0.04 Da. Carbamidomethyl formation was specified as fixed modification, whereas oxidation and acetylation at protein N-terminus were specified as variable modifications and searched against the mouse protein database (Taxonomy ID: 10090) from UniProt (59’783 entries; downloaded: 02.09.2016) concatenated to a decoy (reversed) database and 260 known mass spectrometry contaminants. A target-decoy approach was used to estimate the false-discovery levels (Käll et al. 2008). Peptide and protein assignments were filtered for peptide identification with maximally 5 % false discovery rate (FDR) and protein identification with less than 10 % FDR by Scaffold (Vers. 4.8.3; Proteome Software; Portland, Oregon, USA). The Scaffold spectrum report was exported and loaded into ProgenesisQI. Decoy hits and proteins with single hit peptides were excluded from further analysis. Relative quantification was performed by the Hi-3 approach (Silva et al. 2006). The vitreous proteome of hypoxic wild-type, rodΔVhl and coneΔVhl mice were quantified relatively to their respective age-matched controls. Foldchange (FC) was calculated and statistical significance was determined by one-way ANOVA on the hyperbolic arcsine transformed normalized protein abundance. Significant differential regulation was defined, as |log2(FC)| > 0.58 (FC < 0.67 or FC > 1.5) with a P value < 0.05.
Schori C, Trachsel C, Grossmann J, Barben M, Klee K, Storti F, Samardzija M, Grimm C. A chronic hypoxic response in photoreceptors alters the vitreous proteome in mice. Exp Eye Res. 2019:107690 PubMed: 31181196