The authors have declared that no competing interests exist.
Conceived and designed the experiments: AM FLS. Performed the experiments: AM IP BOB. Analyzed the data: AM IP BOB SB FLS. Contributed reagents/materials/analysis tools: AM SB FLS. Wrote the paper: AM FLS.
In the conjunctiva, repeated or prolonged exposure to injury leads to tissue remodeling and fibrosis associated with dryness, lost of corneal transparency and defect of ocular function. At the site of injury, fibroblasts (FB) migrate and differentiate into myofibroblasts (myoFB), contributing to the healing process together with other cell types, cytokines and growth factors. While the physiological deletion of MyoFB is necessary to successfully end the healing process, myoFB prolonged survival characterizes the pathological process of fibrosis. The reason for myoFB persistence is poorly understood. Nerve Growth Factor (NGF), often increased in inflamed stromal conjunctiva, may represent an important molecule both in many inflammatory processes characterized by tissue remodeling and in promoting wound-healing and well-balanced repair in humans. NGF effects are mediated by the specific expression of the NGF neurotrophic tyrosine kinase receptor type 1 (trkANGFR) and/or the pan-neurotrophin glycoprotein receptor (p75NTR). Therefore, a conjunctival myoFB model (TGFβ1-induced myoFB) was developed and characterized for cell viability/proliferation as well as αSMA, p75NTR and trkANGFR expression. MyoFB were exposed to acute and chronic NGF treatment and examined for their p75NTR/trkANGFR, αSMA/TGFβ1 expression, and apoptosis. Both NGF treatments significantly increased the expression of p75NTR, associated with a deregulation of both αSMA/TGFβ1 genes. Acute and chronic NGF exposures induced apoptosis in p75NTR expressing myoFB, an effect counteracted by the specific trkANGFR and/or p75NTR inhibitors. Focused single p75NTR and double trkANGFR/p75NTR knocking-down experiments highlighted the role of p75NTR in NGF-induced apoptosis. Our current data indicate that NGF is able to trigger
Acute and chronic inflammation of the conjunctiva causes the alteration of local architecture with tissue remodeling and fibrosis associated with ocular dryness and corneal complications leading to visual function impairment
Several findings suggest the contribution of either endogenously produced or topically applied Nerve Growth Factor (NGF) in healing processes, wound-narrowing, tissue remodeling and fibrosis processes
To test whether NGF exposure might modulate myoFB behavior, we induced in vitro the myoFB phenotype via conjunctival FB exposure to TGFβ1. Here we describe evidence that conjunctival TGFβ1-induced myoFB express p75NTR and are more sensitive to apoptosis after NGF treatment.
Conjunctival FB were exposed to 2 ng/mL TGFβ1 for 3 days to develop the myoFB phenotype, according to previous studies
Confluent myoFB were exposed to NGF at the indicated doses/times and evaluated for biochemical and molecular changes. (A) αSMA protein expression in both myoFB and FB, as detected by cell surface ELISA (optic density) and flow cytometry (fluorescence intensity). (B–C) Fluorescent histograms showing a different pattern of trkANGFR and p75NTR expression in acute and chronic 100 ng/mL NGF exposed myoFB, as compared to untreated myoFB (black line). Mean Fluorescent Intensity (MFI) plots are shown inside fluorescent histograms (iso stands for isotype-matched control FI; 7.05 for trkANGFR and 8.39 for p75NTR). (D) Morphological distribution of trkANGFR/p75NTR in untreated, acute and chronic 100 ng/mL NGF treated myoFB. (E–F) Target gene expression specific for trkANGFR/p75NTR after acute and for αSMA/p75NTR after chronic exposure to increasing NGF (p<.05).
When the myoFB were exposed to NGF (100 ng/mL) and harvested 24 hrs from the last stimulation, an increase of p75NTR protein (Mean Fluorescent Intensity, MFI) was detected in both acute (ΔMFI = 90.25, range 57.00 to 173.20;
p75NTR over-expression was associated with αSMA protein (data not shown) and mRNA down-regulation, especially after chronic NGF exposure (
Since a decrease in the number of cells and viability were detected after acute and chronic NGF exposure, NGF exposed monolayers were stained with HO342 and/or DAPI and observed by fluorescence microscopy. Cells with clear signs of nuclear fragmentation were visible as soon as 5 hrs after acute NGF exposure. As shown, acute (
Cells were exposed to different concentrations of NGF and stained with fluorescent DAPI or colorimetric TUNEL. (A-B) An increase of fluorescent cells showing nuclear condensation, picnotic nuclei and perimembrane vescicles is noticeable after acute and chronic NGF treatments. (C) TUNEL images from acute and chronic NGF exposed monolayers. (D) Histogram showing the % of TUNEL-positive NGF treated myoFB, over untreated ones (p<.05).
TUNEL-reactivity confirmed the presence of
Additional studies were performed by using AnnexinV/caspase staining. Outer membrane exposure of phosphatidylserine (PS) was detected by incubating unfixed cells with FC-conjugated AnnexinV. As shown, AnnexinV positivity increased in myoFB after acute (
(A–B) Fluorescent histograms of unsorted (total) cells exposed to acute and chronic NGF doses up to 24 hrs, showing an increased % of AnnexinV positive cells (isotype-matched control FI = 7.65). (C) AnnexinV specific Mean Fluorescent Intensity (MFI) for both acute and chronic, as a function of NGF treatments (% over control untreated myoFB, p<.05). (D) Representative pictures depicting monolayers exposed to acute or chronic 100 ng/mL NGF showing perinuclear/nuclear AnnexinV positivity alone (arrow-heads) or in combination with Propidium Iodide (PI; arrows) in stained cells. (E–F) Cytograms showing AnnexinV/PI plotter in cells exposed to acute and chronic exposure to NGF. Untreated myoFB were mainly AnnexinV/PI double-negative (LL, Lower Left quadrant), with a % range between 73 and 85; both acute and chronic NGF treatments resulted in increases of both AnnexinVpos/PIneg (LR, Lower Right quadrant), indicating a proportion of cells undergoing apoptosis. Particularly, upon chronic NGF treatment, a population of cells also progressed to a later stage of apoptosis (AnnexinVpos/PIpos; UR, Upper Right quadrant).
As observed by studies of double staining, p75NTR expressing myoFB showed round condensed chromatin (picnotic) nuclei. This TUNEL/p75NTR coexpression was particularly evident for 100 ng/mL NGF treated myoFB, in comparison to TGFβ1 treated ones (
(A) Pictures depicting TUNEL reactivity in p75NTR positive cells exposed to chronic 100 ng/mL NGF, as compared to TGFβ1 exposed cells. Changes in morphology, such as having rounded cytoplasm and markedly condensed chromatin/nuclear fragmentation, are clear visible in p75NTR-bearing myoFB. A few TUNEL-positive myoFB were also quantified in untreated cultures (not shown). (B) Overlays and single staining specific for p75NTR and cleaved caspase3 are visible in both acute and chronic 100 ng/mL NGF exposed myoFB, as compared to untreated myoFB. (C–D) p75NTR positive sorted cells analysed for active caspase3 expression and related statistical analysis of MFI. Note the specific reduction upon pretreated with trkANGFR and/or p75NTR chemical inhibitors.
In these p75NTR bearing cells, the contribution of caspase pathway was also investigated by using terminal effector caspase3 antibodies. Overlays and single staining specific for cleaved (active) caspase3 or p75NTR expression are shown in untreated, acute and chronic NGF exposed myoFB (
Since myoFB express both trkANGFR/p75NTR on their surface, NGF-induced apoptosis was also investigated in the presence of trkANGFR and p75NTR specific inhibitors, alone or in combination. As noted, active caspase3 signal was significantly reduced after preincubation with trkANGFR and/or p75NTR chemical inhibitors (500 ng/mL;
To provide further information on p75NTR and particularly trkANGFR/p75NTR contribution in NGF-mediated apoptosis, appropriate silencing (siRNA) experiments were designed (see flow chart in
(A) A flow chart schematizing the siRNA experiments. (B) Confocal images depicting GFP positive cells (image) and p75NTR-GFP double-positive cells (insert over the image), re-plated 5 hrs after electroporation. (C–D) Representative histogram for GFP specific staining, in comparison to control siRNA (control cells), showing a decreased p75NTR positive expression and the related p75NTR target gene deregulation. (E) Mean Fluorescence Intensity (MFI) specific for caspACE-VAD comparing 100 ng/mL NGF and 10 ng/mL TGFβ1 effects on control-siRNA, p75NTR siRNA and double trkANGFR/p75NTR siRNA treated cells. A 37.5%-decrease (*) and a 50%-decrease (**) of apoptotic signal were detected respectively in p75NTR-siRNA and p75NTR/trkANGFR-siRNA (p<.05).
The main finding of this study indicates that TGFβ1-induced conjunctival myoFB (herein referred as myoFBs) significantly increase their p75NTR and undergo apoptosis upon acute and particularly chronic NGF treatment. Both chemical inhibition and single- (p75NTR) and double (trkANGFR/p75NTR)-silencing approaches (siRNA) down-regulated the apoptotic signal, highlighting the contribution of NGF/p75NTR in mediating myoFB apoptosis.
Tissue remodeling and fibrosis clearly compromise ocular surface structure and lead to ocular function decline
In order to investigate the NGF effect on myoFB, a cell culture model of TGFβ1-induced myoFB was reproduced and firstly characterized for NGF receptor pathway. TGFβ1 is a widely accepted chief inducer of FB differentiation, even if overlapping mechanisms and soluble factors might contribute massively to the development of myoFB phenotype in inflamed tissues,
Next, the NGF receptor pathway was investigated in stimulating and neutralizing experiments. Acute and massively chronic NGF exposure triggered a selective increase of p75NTR expression while trkANGFR expression was slightly modulated. Both treatments shifted the trkANGFR/p75NTR rate toward p75NTR expression, in a dose- and time-dependent manner, resulting in a decrease of the trkANGFR/p75NTR ratio. As previously reported, p75NTR expression is a typical feature of healing myoFB
As previously reported, conditions leading to a lower trkANGFR/p75NTR ratio are often associated with apoptosis
Arising from quiescent FB and transiently expressed during wound-healing, myoFB play a crucial role in the late phases of the repair process being responsible for the ECM deposition and contraction
In our study, the cells showing DNA fragmentation were also recognized as p75NTR bearing cells, as detected by sorting analysis. NGF exposure resulted in increasing cleavage of caspase-3 significantly expressed in p75NTR bearing myoFB. Even though in this cell culture model, the apoptotic pathway induced by NGF appears associated with activation of the effector caspase-3, the intracellular death-pathway induced by p75NTR activation is unknown.
According to the literature, trkANGFR mainly drives proliferation, differentiation and survival while p75NTR mainly triggers differentiation and apoptosis
Therefore to provide additional information, apoptosis was also investigated in NGF-treated myoFB expressing low trkANGFR/p75NTR ratio (100 ng/mL) by using small inhibitory RNAs (siRNA) directed against p75NTR
Taken together, our in vitro model provides data on NGF as an inducer of apoptosis by shifting the trkANGFR/p75NTR ratio in favor of p75NTR and inducing up-regulation of terminal caspase-3, opening a more fruitful possibility for counteracting myoFB resistance to apoptosis. Indeed, the observed NGF-mediated TGFβ1 deregulation is in line with the apoptotic effect and with the inhibition of TGFβ1-mediated myoFB endurance. In line, our data might explain the observation of the successful corneal healing upon NGF topical application
Purified 2.5S-NGF (GradeI) and purified human TGFβ1 were provided by Alomone Labs Ltd. (Jerusalem, Israel) and PeproTech EC Ltd (London, UK) respectively; neutralizing anti-pan TGFβ antibodies and neutralizing anti-NGF antibodies (500 ng/mL) were from R&D Systems, Minneapolis, MN; neutralizing anti-trkANGFR (100 ng/mL) and anti-p75NTR (100 ng/mL) were from Calbiochem Novabiochem Corp. (San Diego, USA). Conjugated specie-specific secondary antibodies were from Jackson ImmunoResearch Europe Ltd (Suffolk, UK). Other analytical grade chemicals and solvents were from ICN (Costa Mesa, CA), SERVA (Weidelberg, Germany) and Sigma Chemicals (St Louis, Mo), unless otherwise indicated. All tissue culture reagents were from Lonza (Basel, Switzerland) and sterile tissue culture plastic-ware were from NUNC (Roskilde, Denmark). Grade reagent and Ultrapure/RNAse free water was provided by Direct-Q 5 Apparatus (Millipore, Vimodrone, Milan, Italy).
The research protocol followed the tenets of the Declaration of Helsinki and was approved by the Intramural Committee of Hadassah University Hospital. Human primary cultures of conjunctival FB (passages 3–7) were obtained from Innoprot (Bizkaia, Spain) and used for these studies. Cell cultures were expanded (24 hrs doubling time) in DMEM supplemented with 10% heat-inactivated Fetal Calf Serum (FCS), 2 mM-glutamine, 100 U/ml pen-streptomycin, under standard culturing conditions and re-plated after trypsin-EDTA harvesting. To obtain
TGFβ1-induced myoFB (herein termed myoFB and defined as αSMA-expressing myoFB)
Cells were seeded on 96-well plates and then treated with increasing NGF doses and times depending on the experiment. Cell viability was determined by measuring mitochondrial reduction of the MTS dye reagent into a soluble formazan product, according to the manufacturer’s instructions (Promega, Madison, WI). Cell surface ELISA was carried out for quantifying αSMA expression, as previously reported in detail
Single cells (106cells/well) were processed for membrane/cytoplasm staining (PE-trkANGFR/FC-p75NTR diluted 1/50; Santa Cruz Biotec, Santa Cruz, CA), according to a standardized protocol including mild postfixation (0.3% ρ-formaldehyde (PFA)), brief methanol permeabilization, blocking (5% FCS) treatments and staining in PBS containing 5% FCS, 1 mM EDTA and 0.05% NaN3 (FACS buffer). Cells were incubated with mouse anti-human αSMA antibodies (sc-130616) followed by species-specific APC-conjugated anti-Mouse IgG antibodies; or probed with a PE-conjugated anti-human trkANGFR antibodies (sc-118) and FC-conjugated anti-human p75NTR antibodies (sc-81612) mixture (Santa Cruz Biotech., Santa Cruz, CA). Cleaved caspase-3 antibodies (Asp.175 #9661; Cell Signaling Technology Inc., Danvers, MA) were labeled with PerCP specie-specific antibodies. Cells were evaluated using a digital based flow cytometry station (MACSQuant Analyzer, Miltenyi). Hyperlog (hLog) signals were analysed from 5000 gated cells/sample. Instrument calibration was checked weekly by use of Microbeads (Miltenyi) and individual compensation settings for each separate reagent combination (tube-specific compensations) were performed by use of antibody-capture beads (CompBeads, Becton Dickinson, San Jose, CA). Instrument settings were performed using control (isotype-matched antibodies from eBiosciences, San Diego, CA), single and double fluorescent samples, run in parallel for each set of experiments. MFI of hLog distribution was calculated, non-specific signal for each sample was subtracted from the specific one, and results were expressed as increments relative to the controls, calculated as follows: ΔMFI = (specific MFI - non-specific MFI)/non-specific MFI. A MFI ratio>1 represents significant expression. Histogram or density plots were arranged using the MACSQuant Digital software. MFI data are expressed as means±SD.
Cells on round coverslips (2x105) were washed in 10 mM Phosphate-Buffered Saline (PBS, pH7.5) at the stated time-points, fixed in 3.7% buffered PFA, quenched in 50 mM NH4Cl to rule out autofluorescence and blocked/permeabilized in 3% BSA and 0.03% Triton X-100 (TX) in PBS. Monolayers were probed with: rabbit anti- trkANGFR antibodies and goat anti- p75NTR antibodies (4 µg/mL, both from R&D); mouse anti-αSMA (2 µg/mL; Santa Cruz) and active caspase3 (Cell signaling). Specific binding of the primary antibody was detected using Cy2 or APC-conjugated secondary antibodies (1/500). Nuclear counter-staining was performed with Propidium Iodide (25 µg/mL) or with TOTO-3 (1 mM) in PBS containing RNase (20 µg/mL) and coverslips were mounted in hand-made anti-fade medium. Irrelevant isotype-matched IgG antibodies (Vector) were incubated in parallel and used as internal controls for the channel series acquisitions and related background subtraction. Monolayers were examined and images were acquired using an inverted E2000-U microscope equipped with the C1 software (x20/0.45NA; x40/0.60NA; x60/1.4 oil; Nikon, Tokyo, Japan). TIFF-converted pictures were assembled by Adobe Photoshop 7.0 (Adobe Systems Inc., San Jose, CA).
Total RNA was extracted from cells (2×105) treated and harvested at 5 or 24 hrs from last stimulation, using the EuroGold TRIfast™ kit (Euroclone, Milan, Italy) and diluted in RNase free water (Millipore). The concentration and purity (260/280 nm and 260/230 nm) of total RNA were determined by using a spectrophotometer (NanoDrop® ND-1000, Wilmington, DE, USA), while 28S/18S ratios were recorded after agarose gel separation. When required, a DNAse treatment was allowed to guarantee absence of contaminating DNA (Turbo DNA free kit; Ambion, Milan, Italy). Only RNAs showing RNA/DNA rate >1.8 were used for analysis. Real time PCR was performed in a two step manner: cDNA synthesis and amplification were carried out in a One Personal thermocycler (PeqLab, EuroClone) and in an Opticon2 system (MJ Research, Watertown, MA), respectively. Oligo dT21-primer was used to generate cDNA from 1 µg total RNA, according to the manufacturer’s instructions (Improm kit; Promega). SYBRGreen HotStart AmpliGold Taq polymerase (Applied Biosystems, Foster City, CA) was used for specific amplification of 3 µL cDNA in a final volume of 20 µL in 96-well plate including internal controls, at stated amplification settings (see
Sequence | Amplicon (bps) | Annealing conditions |
|
F: |
100 bps | 53°C, 30 sec | |
F: 5′-TCC TGG CGA TAC CTC AGC AA-3′R: 5′-GCC CTC AAT TTC CCC TCC AC-3′ | 110 bps | 53°C, 30 sec | |
F: 5′-GAA GGA GAT CAC GGC CCT A-3′R: 5′- ACA TCT GCT GGA AGG TGG AC-3′ | 125 bps | 60°C, 25 sec | |
F: 5′- CTG GCC ACA CTG AGG TGC AT-3′R: |
120 bps | 53°C, 30 sec | |
F: |
103 bps | 57°C, 25 sec | |
F: |
147 bps | 57°C, 25 sec | |
|
|
||
S: 5′- CCGUUUGCUGUGAACCCUAUGUUAU -3′AS: 5′- AUAACAUAGGGUUCACAGCAAACGG -3′ | |||
S: 5′- CCAUCGUGAAGAGUGGUCUCCGUUU -3′AS: 5′- AAACGGAGACCACUCUUCACGAUG G-3′ |
A summary of primer names, for/rev primer sequences (5′ to 3′), PCR product size (amplicons in bps), annealing conditions and Genebank accession numbers of each gene investigated. F: forward primer; R: reverse primer.
PCR primers were designed by Primer3 software (genome.wi.mit.edu/cgi-bin/primer/primer3_
The RNAi were designed by Invitrogen website (rnaidesigner.invitrogen.com/).
Amplification parameters were as follows: 37 cycles of 30 s/94°C, 25s/specific Ta, 30 s/72°C, preceded by 15 m/95°C hot start polymerase activation and followed by fluorescence monitoring during linear transition from 55–90°C, 0.01°C for 0.3 sec and further 5 m/72°C incubation.
Conditioned media were gently removed and centrifuged to collect death cells, and adherent cells were briefly trypsin harvested and processed as reported in detail.
Cells (2x105) were treated with NGF for 5 hrs and stained with 1 µg/mL Hoechst 33342 (HO342) for 5 min at 37°C or 1 µg/mL DAPI for 15 min at 37°C (both from Invitrogen-Molecular Probes, Milan, Italy) to identify cells with membrane shrinkage, chromatin condensation and clumping or enlarged cells. Monolayers were photographed with a microscope equipped with epifluorescence to detect Hoechst/DAPI signals and software for image acquisition (Nikon, Tokyo, Japan). TIFF-converted pictures were assembled by Adobe Photoshop 7.0.
Adherent cells (2×105) were stained with the biotinyled dATP nick end labeling procedure (TUNEL), which mark DNA fragmentation. Briefly, fixed (3.7% PFA) and permeabilized (0.1% TX) cells were incubated with exogenous rTdT enzyme and biotin 14-dATP (37°C/3 hrs; Invitrogen), for repair of 3′-hydroxyl DNA ends. Positive controls were carried out in parallel, with a DNase I pre-incubation of monolayers (2 U/mL; 37°C/30 min). Negative controls were monolayers that were incubated with buffer lacking rTdT enzyme. The apoptotic nuclei were labeled according to the ABC technique (Vector Laboratories, Burlingame, CA) using DAB solution (Dako Corp., Carpinteria, CA) as substrate. TUNEL positive cells were scored by two independent observers (blind fashion), and five optic fields/coverslip for at least 10 coverslips were counted for each culture condition (x20). The percentage of apoptosis was calculated as (number of apoptotic cells/number of total cells)×100. For double staining with p75NTR and specific FC-secondary antibodies, TUNEL reaction was performed according to the TRITC-Avidin procedure (Vector). Stained cells were imaged as reported in the confocal laser microscopy section.
Single harvested cells (106) were equilibrated in Hepes buffer containing 2 mM CaCl2, and subjected to immunofluorescence as follows. To detect externalized PS, cells were probed with FC-AnnexinV-Apoptosis detection kit (Miltenyi), according to the manufacturer’s directions with the exception that cells were post treated with RNAse A and finally fixed in 3% PFA
Specific Stealth™ siRNA molecules were chemically synthesized, as desalted, 25mer duplex oligonucleotides (
All the experiments were done independently at least three times, each experiment being carried out in duplicate or triplicate. Data (means±SD) were exported to Excel datasheet and presented as graphs (Excel Microsoft Corp., Redmond, WA; Microsoft.com/excel). Statistical analyses were performed by using the StatView II package for PC (Abacus Concepts. Inc., Berkeley, CA). All data followed a normal distribution and parametric ANOVA followed by Tukey-Kramer post hoc was employed to analyze data. The Spearman’s rank correlation coefficient (Rho) was calculated to identify specific target relationships. A confidence of 95% was presumed to reflect significant difference between group mean values, and therefore a p-value of <.05 was taken as the limit of significance. A specific REST/ANOVA coupled analysis was carried out for PCR experiments.
We are indebted to Prof. Rita Levi-Montalcini for her stimulating discussions and suggestions in our research that was the main part of the PhD program in Immunopharmacology of A. Micera (at the HUJI). We thank Ramona Marino from the laboratory of Prof. Keller (Dept. Developmental Neurobiology at the UCBM), for her contribution in acquiring fluorescent images.