University of Groningen Identification of biomarkers for diabetic retinopathy Fickweiler, Ward

(1)

University of Groningen

Identification of biomarkers for diabetic retinopathy

Fickweiler, Ward

DOI:

10.33612/diss.95666609

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from

it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date:

2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Fickweiler, W. (2019). Identification of biomarkers for diabetic retinopathy. University of Groningen.

https://doi.org/10.33612/diss.95666609

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

C h ap te r C h ap te r 2. 2 C h ap te r 2. 2 C h ap te r 2. 2 C h ap te r 2. 2

chapter 2.2

retinal proteome associated with

bradykinin-induced edema

Nivetha Murugesan, Ward Fickweiler, Allen C. Clermont, Qunfang Zhou, Edward P. Feener

(3)

Chapter 2.2 abstract

The plasma kallikrein-stimulated generation of bradykinin (BK) has been implicated in dia-betic macular edema (DME). This study characterizes the effects of BK on the ultrastructure and proteome of the rat retina. The effects of intravitreal injection of BK on retinal thickness and vascular ultrastructure in Sprague Dawley rats were analyzed and compared with the ef-fects of VEGF using spectral-domain optical coherence tomography. At 24 h post intravitreal injection of BK or saline vehicle retina were harvested and solubilized proteins were analyzed by mass spectrometry-based proteomics. Proteins were identified using X!Tandem and spec-tral counts were used as a semiquantitative measurement of protein abundance. Proteins identified from retinal extracts were annotated by Gene Ontology (GO) slim terms and com-pared with a human DME vitreous proteome. Intravitreal injection of BK and VEGF induced transient increases in retinal thickness of 46 µm (24.6%, p = 0.015) and 39 µm (20.3%, p = 0.004), respectively at 24 h, which were resolved to baseline thicknesses at 96 h post injec-tion. BK and VEGF also increased retinal vessel diameters and tortuosity at 24 h post intra-vitreal injection. Proteomic analyses identified 1757 non-redundant proteins in the rat retina, including 1739 and 1725 proteins from BK- and saline control-injected eyes, respectively. Eighteen proteins, including two proteins associated with intercellular junctions, filamin A and actinin alpha 4, were decreased by at least 50% (p < 0.05) in retina from BK-injected eyes compare with retina from eyes injected with saline. In addition, 32 proteins were increased by > 2-fold (p < 0.05) in retina from BK-injected eyes. Eight proteins, including complement C3, were identified to be increased in both BK-stimulated rat retina and in human DME vitreous. Western blot analysis showed that Complement 3 levels in vitreous from BK-injected eyes in rats and clinical DME samples were increased by 6.6- fold (p = 0.039) and 4.3-fold (p = 0.02), compared with their respective controls. In summary, this study identifies protein changes in rat retina that are associated with BK-induced retinal thickening, including 8 proteins that were previously reported to be increased in the human DME vitreous proteome.

(4)

39 C h ap te r 2. 2 1. IntroductIon

Diabetic macular edema (DME) is the leading cause of moderate vision loss in working-aged adults of most developed countries (Lee et al., 2015). This vision threatening disease is caused by the breakdown of the blood retinal barrier; resulting in retinal vascular hyperpermeability and the accumulation of fluid in the retina leading to central subfield thickening and loss of visual acuity (Duh et al., 2017). Previous studies by our group and others have shown that the vitreous proteome is markedly altered in DME compared with vitreous from diabetic subjects without diabetic retinopathy (Gao et al., 2008; Kita et al., 2015; Loukovaara et al., 2015). While these reports have characterized the proteomic changes in the DME vitreous much less is known regarding the proteomic changes that occur in the retina during retinal edema. Characterization of the proteome in retinal edema may provide new insight into the mechanisms that contribute to this condition.

Plasma kallikrein (PKa) has been implicated in mediating both vascular endothelial growth factor (VEGF)-dependent and –independent causes of DME (Clermont et al., 2016; Gao et al., 2007; Kita et al., 2015). PKa is a serine protease that is derived from its zymogen plasma prekallikrein (PK), which is abundantly present in the blood. PK is proteolytically activated to PKa by the serine protease FXIIa. Subsequently, PKa cleaves its substrate high molecular weight kininogen (HK) to generate cleaved HK (cHK) and the nonapeptide hormone bra-dykinin (BK). BK activates brabra-dykinin B1 and B2 receptors, which are G-protein coupled receptors expressed in the retina (Abdouh et al., 2003; Kita et al., 2015; Ma et al., 1996). BK can exert pro-inflammatory effects including increased vascular permeability, vasodilation, vascular permeability, pain, and immune cell activation (Liu and Feener, 2013). PKa and the BK receptor, B2R, are clinically significant mediators of edematous attacks in hereditary angioedema (Riedl, 2012).

In previous studies, our group and others have shown that concentrations of kallikrein kinin system (KKS) components, including PK, PKa, FXII, FXIIa, and cHK are increased in the vitreous from DME patients compared with controls with nondiabetic control subjects with macular hole (Gao et al., 2007; Kita et al., 2015). The KKS has been reported to contribute to retinal vascular permeability and edema mediated by diabetes, hypertension, hemorrhage, TNFα, and VEGF (Clermont et al., 2011, 2016; Gao et al., 2007; Liu et al., 2013; Phipps et al., 2009). While DME has been shown to increase levels of the KKS proteins in the vitreous the effects of this system on molecular changes in the retina are not fully understood.

In this report the effects of intravitreal injection of BK on the retinal ultrastructure and proteome in rats are characterized. The effects of intravitreal administration of BK on retinal thickening, vasodilation, and vessel tortuosity were compared with the effects of VEGF, which was used as a positive control (Clermont et al., 2016). Protein changes in the retina associated with BK-induced retinal edema in rats were identified and compared with the protein changes in vitreous samples from DME patients.

(5)

Chapter 2.2

high molecular weight kininogen (HK) to generate cleaved HK (cHK) and the nonapeptide hormone bradykinin (BK). BK activates bradykinin B1 and B2 receptors, which are G-protein coupled receptors expressed in the retina (Abdouh et al., 2003;Kita et al., 2015;Ma et al., 1996). BK can exert pro-inflammatory effects including increased vascular per-meability, vasodilation, vascular perper-meability, pain, and immune cell activation (Liu and Feener, 2013). PKa and the BK receptor, B2R, are clinically significant mediators of edematous attacks in hereditary an-gioedema (Riedl, 2012).

In previous studies, our group and others have shown that con-centrations of kallikrein kinin system (KKS) components, including PK, PKa, FXII, FXIIa, and cHK are increased in the vitreous from DME pa-tients compared with controls with nondiabetic control subjects with macular hole (Gao et al., 2007;Kita et al., 2015). The KKS has been reported to contribute to retinal vascular permeability and edema mediated by diabetes, hypertension, hemorrhage, TNFα, and VEGF (Clermont et al., 2011,2016;Gao et al., 2007;Liu et al., 2013;Phipps et al., 2009). While DME has been shown to increase levels of the KKS proteins in the vitreous the effects of this system on molecular changes in the retina are not fully understood.

In this report the effects of intravitreal injection of BK on the retinal ultrastructure and proteome in rats are characterized. The effects of intravitreal administration of BK on retinal thickening, vasodilation, and vessel tortuosity were compared with the effects of VEGF, which was used as a positive control (Clermont et al., 2016). Protein changes in the retina associated with BK-induced retinal edema in rats were identified and compared with the protein changes in vitreous samples from DME patients.

2. Methods

2.1. Animals

Male Sprague-Dawley rats at 8 weeks of age (230–275 g) were purchased from Envigo (Indianapolis, IN). Intramuscular injection of ketamine (80 mg/kg; VEDCO, St. Joseph, MO) and xylazine (10 mg/kg; Sigma-Aldrich, Milwaukee, WI) were used as anesthesia for rats. All experiments were performed in accordance with guidelines from the Association for Research in Vision and Ophthalmology and with ap-proval from the Institutional Animal Care and Use Committee of the Joslin Diabetes Center.

2.2. Intravitreal administration of bradykinin and VEGF

Intravitreal (IVT) injections in rats were performed as described previously (Clermont et al., 2016;Kita et al., 2015). Briefly, rats were anesthetized, eyes were dilated with 1% tropicamide and received a single intravitreal injection of 5 μL of Phosphate Buffered Saline (PBS) as vehicle (Alcon Laboratories, Fort Worth, Texas, USA), 2 μM BK (EMD Millipore, Billerica, MA), or human recombinant VEGF (10 ng/eye, Life Technologies/Gibco, Grand Island, NY) using a 10 μL Hamilton syringe with a 31-gauge needle.

2.3. Optical coherence tomography

Retinal thickness in rats was measured by Spectral domain–optical coherence tomography (SD-OCT) using an 840SD OCT System (Bioptigen, Durham, NC). Briefly, rectangular volumes of retina were obtained from 1000 A-scans by 100 B-scans over a 2.5 × 2.5 mm retinal area centered upon the optic nerve head (ONH). Retinal layers were measured at 600 μm relative to the ONH at eight points defined by calipers on the OCT derived en face image as shown inFig. 1A, and as

Fig. 1. Effects of Bradykinin and VEGF on retinal vessel diameters and tortuosity at 24 h post intravitreal injection in rats. (A) Representative spectral-domain optical

coherence tomography (SD-OCT) derived en face images at baseline and 24 h post intravitreal injection of PBS, BK and VEGF. Retinal layers were measured at 600 μm relative to the ONH at eight points defined by calipers on the OCT-derived en face image (the approximate positions of calipers are shown in PBS-Baseline image). (B) Retinal vessel diameters at baseline (pre-injection) and 24 h post intravitreal injection of BK and VEGF. (C) Retinal tortuosity at baseline (pre-injection) and 24 h after intravitreal injection (post-injection) of BK and VEGF. *p < 0.05 using t-test, n = 6 animals per group.

N. Murugesan, et al. Experimental Eye Research 186 (2019) 107744

2

fig. 1. Effects of Bradykinin and VEGF on retinal vessel diameters and tortuosity at 24 h post intravitreal

injection in rats. (a) Representative spectral-domain optical coherence tomography (SD-OCT) derived en face images at baseline and 24 h post intravitreal injection of PBS, BK and VEGF. Retinal layers were measured at 600 µm relative to the ONH at eight points defined by calipers on the OCT-derived en face image (the approximate positions of calipers are shown in PBS-Baseline image). (b) Retinal vessel diam-eters at baseline (pre-injection) and 24 h post intravitreal injection of BK and VEGF. (c) Retinal tortuosity at baseline (pre-injection) and 24 h after intravitreal injection (post-injection) of BK and VEGF. *p < 0.05 using t-test, n = 6 animals per group.

2. methods 2.1. Animals

Male Sprague-Dawley rats at 8 weeks of age (230–275 g) were purchased from Envigo (In-dianapolis, IN). Intramuscular injection of ketamine (80 mg/kg; VEDCO, St. Joseph, MO) and xylazine (10 mg/kg; Sigma-Aldrich, Milwaukee, WI) were used as anesthesia for rats. All experiments were performed in accordance with guidelines from the Association for Research in Vision and Ophthalmology and with approval from the Institutional Animal Care and Use Committee of the Joslin Diabetes Center.

2.2. Intravitreal administration of bradykinin and VEGF

Intravitreal (IVT) injections in rats were performed as described previously (Clermont et al., 2016; Kita et al., 2015). Briefly, rats were anesthetized, eyes were dilated with 1% tropicamide and received a single intravitreal injection of 5 µL of Phosphate Buffered Saline (PBS) as vehicle (Alcon Laboratories, Fort Worth, Texas, USA), 2 µM BK (EMD Millipore, Billerica, MA), or human recombinant VEGF (10 ng/eye, Life Technologies/Gibco, Grand Island, NY) using a 10 µL Hamilton syringe with a 31-gauge needle.

(6)

41 C h ap te r 2. 2

2.3. Optical coherence tomography

Retinal thickness in rats was measured by Spectral domain–optical coherence tomography (SD-OCT) using an 840SD OCT System (Bioptigen, Durham, NC). Briefly, rectangular vol-umes of retina were obtained from 1000 A-scans by 100 B-scans over a 2.5 × 2.5 mm retinal area centered upon the optic nerve head (ONH). Retinal layers were measured at 600 µm relative to the ONH at eight points defined by calipers on the OCT derived en face image as shown in Fig. 1A, and as described previously (Clermont et al., 2011). At the intersection of the distance calipers and the corresponding B-scan, retinal layers were measured with calibrated calipers using Bioptigen InVivoVue version 1.4. The measurements were averaged to produce a single thickness value for each retinal layer. Retinal vessel diameter and tortuos-ity were measured using calibrated calipers on the OCT-derived en face image. Tortuostortuos-ity was defined as the maximum angle of a vessel > 200 µm from the ONH. The maximum angle was subtracted from a straight line (180°) to determine the tortuosity of a vessel. Caliper measurements were taken for each vessel and repeated five times and were used to compare the same vessels at baseline and at 24 h post injection. Measurements of the three largest vessels were averaged to produce a single value for each image.

described previously (Clermont et al., 2011). At the intersection of the distance calipers and the corresponding B-scan, retinal layers were measured with calibrated calipers using Bioptigen InVivoVue version 1.4. The measurements were averaged to produce a single thickness value for each retinal layer. Retinal vessel diameter and tortuosity were measured using calibrated calipers on the OCT-derived en face image. Tortuosity was defined as the maximum angle of a vessel > 200 μm from the ONH. The maximum angle was subtracted from a straight line (180°) to determine the tortuosity of a vessel. Caliper measurements were taken for each vessel and repeated five times and were used to compare the same vessels at baseline and at 24 h post injection. Mea-surements of the three largest vessels were averaged to produce a single value for each image.

2.4. Retinal proteomics

Rat retinal samples (n = 4 rats for BK-injected eyes, n = 3 rats for PBS-injected eyes) were harvested 24 h after IVT injection, lysed in

TPER tissue extraction reagent (Thermo Scientific, Rockford, IL), and solubilized proteins were separated by SDS-PAGE. Proteins were stained with Coomassie Brilliant Blue (Bio-Rad, Hercules, CA, USA), gel lanes were cut into 40 equal slices, and tryptic digests (Sequencing grade modified trypsin, porcine, Promega, Madison, WI) were prepared from each slice, as described previously (Gao et al., 2009). Tryptic di-gests of each rat retina were analyzed individually by nanospray liquid chromatography tandem mass spectrometry (MS/MS) using a LTQ linear ion trap mass spectrometer (ThermoFisher, Waltham, MA). MS/ MS spectra were analyzed using X!Tandem (The Global Proteome Ma-chine Organization,thegpm.org; version X! Tandem) and the rat se-quence database uni.ReverseRatConcat.2015_03.fasta.pro database. Scaffold 4.4.1.1 (Proteome Software Inc., Portland, OR) was used to analyze and compile MS/MS based peptide and protein identifications. X!Tandem search parameters included the following: maximum valid expectation value of 0.1; residue mass modifications of +16.0 Da for oxidized methionine and +71.0 Da for acrylamide alkylated cysteine; fragment monoisotopic mass error of ± 0.4 Da and precursor mono-isotopic mass error of ± 0.5 Da. Peptide identifications were accepted if they could be established at greater than 90.0% probability with 0.01% false discovery rate (FDR) by the Scaffold Local FDR algorithm. Search results were compiled into a MySQL database and analyzed using MS Results Manager as described previously (Gao et al., 2008). Protein identifications were accepted if they could be established at greater than 99.0% probability and contained at least 2 identified peptides with 0.1% FDR in the same or adjacent gel slices. Protein probabilities were assigned by the Protein Prophet algorithm (Nesvizhskii et al., 2003) and proteins were annotated with Gene Ontology (GO) (Ashburner et al., 2000) terms from NCBI using DAVID Bioinformatics Resources 6.8, NIAID/NIH.

2.5. Western blotting of vitreous and retinal lysates

Vitreous samples (n = 4 per group) were obtained from patients during surgery for macular hole (MH) or DME with nonproliferative diabetic retinopathy, as described previously (Kita et al., 2015). Rat retinal extracts (n = 5 per group) were prepared from perfused retina using TPER (Thermo Scientific, Waltham, MA) with protease inhibitor cocktail (Sigma, Milwaukee, WI). Western blotting was performed using rabbit anti-C3 antibody (Abcam, Cambridge, MA). The results were then visualized by enhanced chemiluminescence (20x Lumiglo, Cell Signaling, Danvers, MA) and quantified using ImageJ software (NIH, Bethesda, MD).

2.6. Statistics

All results were expressed as mean ± standard error of the mean (SEM). The statistical significance of differences between groups was analyzed by one way analysis of variance (ANOVA), with Holm-Sidak repeated measures multiple comparisons test for the retinal thickness measurements over time. ANOVA with Dunnett's post hoc test was performed for the segmental OCT thickness measurements (SigmaPlot V12, San Jose, California). Differences were considered significant at p < 0.05. For the proteomic studies, the correlation of spectral counts and retinal thickness change at 24 h post IVT injection were analyzed using the Pearson correlation coefficient r.

3. Results

3.1. Effects of Bradykinin and VEGF on retinal ultrastructure

We examined and compared the effects of intravitreal injections of BK, VEGF, and saline control on retinal vessel diameter and tortuosity in rats using SD-OCT. Representative en face retinal images at baseline and at 24 h post injection show retinal vascular dilation and increased tortuosity in eyes subjected to BK and VEGF injections compared with

Fig. 2. Time course of Bradykinin- and VEGF-induced retinal thickening in rats

(A) The effect of intravitreal injection of VEGF, BK and BK/VEGF co-injection on retinal thickness over a time course of 6 h to 5 days using SD-OCT. *p < 0.05, **p < 0.005 compared to baseline using one-way ANOVA with Holm-Sidak repeated measures multiple comparisons test, n = 7 animals per group. (B) Layer thickness measured in each retinal segment of PBS (n = 6), BK (n = 6) and VEGF (n = 6) injected retina using calibrated calipers. The mea-surements were averaged to produce a single thickness value for each retinal layer. NFL: Nerve Fiber Layer, IPL: inner plexiform layer, INL: inner nuclear layer, ONL: outer nuclear layer, SEG: photoreceptor segments, *p < 0.05, **p < 0.005 compared to baseline to 24 h post injection using one-way ANOVA with Dunnett's post hoc test, n = 7 animals per group.

3 described previously (Clermont et al., 2011). At the intersection of the

distance calipers and the corresponding B-scan, retinal layers were measured with calibrated calipers using Bioptigen InVivoVue version 1.4. The measurements were averaged to produce a single thickness value for each retinal layer. Retinal vessel diameter and tortuosity were measured using calibrated calipers on the OCT-derived en face image. Tortuosity was defined as the maximum angle of a vessel > 200 μm from the ONH. The maximum angle was subtracted from a straight line (180°) to determine the tortuosity of a vessel. Caliper measurements were taken for each vessel and repeated five times and were used to compare the same vessels at baseline and at 24 h post injection. Mea-surements of the three largest vessels were averaged to produce a single value for each image.

Rat retinal samples (n = 4 rats for BK-injected eyes, n = 3 rats for PBS-injected eyes) were harvested 24 h after IVT injection, lysed in

TPER tissue extraction reagent (Thermo Scientific, Rockford, IL), and solubilized proteins were separated by SDS-PAGE. Proteins were stained with Coomassie Brilliant Blue (Bio-Rad, Hercules, CA, USA), gel lanes were cut into 40 equal slices, and tryptic digests (Sequencing grade modified trypsin, porcine, Promega, Madison, WI) were prepared from each slice, as described previously (Gao et al., 2009). Tryptic di-gests of each rat retina were analyzed individually by nanospray liquid chromatography tandem mass spectrometry (MS/MS) using a LTQ linear ion trap mass spectrometer (ThermoFisher, Waltham, MA). MS/ MS spectra were analyzed using X!Tandem (The Global Proteome Ma-chine Organization,thegpm.org; version X! Tandem) and the rat se-quence database uni.ReverseRatConcat.2015_03.fasta.pro database. Scaffold 4.4.1.1 (Proteome Software Inc., Portland, OR) was used to analyze and compile MS/MS based peptide and protein identifications. X!Tandem search parameters included the following: maximum valid expectation value of 0.1; residue mass modifications of +16.0 Da for oxidized methionine and +71.0 Da for acrylamide alkylated cysteine; fragment monoisotopic mass error of ± 0.4 Da and precursor mono-isotopic mass error of ± 0.5 Da. Peptide identifications were accepted if they could be established at greater than 90.0% probability with 0.01% false discovery rate (FDR) by the Scaffold Local FDR algorithm. Search results were compiled into a MySQL database and analyzed using MS Results Manager as described previously (Gao et al., 2008). Protein identifications were accepted if they could be established at greater than 99.0% probability and contained at least 2 identified peptides with 0.1% FDR in the same or adjacent gel slices. Protein probabilities were assigned by the Protein Prophet algorithm (Nesvizhskii et al., 2003) and proteins were annotated with Gene Ontology (GO) (Ashburner et al., 2000) terms from NCBI using DAVID Bioinformatics Resources 6.8, NIAID/NIH.

Vitreous samples (n = 4 per group) were obtained from patients during surgery for macular hole (MH) or DME with nonproliferative diabetic retinopathy, as described previously (Kita et al., 2015). Rat retinal extracts (n = 5 per group) were prepared from perfused retina using TPER (Thermo Scientific, Waltham, MA) with protease inhibitor cocktail (Sigma, Milwaukee, WI). Western blotting was performed using rabbit anti-C3 antibody (Abcam, Cambridge, MA). The results were then visualized by enhanced chemiluminescence (20x Lumiglo, Cell Signaling, Danvers, MA) and quantified using ImageJ software (NIH, Bethesda, MD).

2.6. Statistics

All results were expressed as mean ± standard error of the mean (SEM). The statistical significance of differences between groups was analyzed by one way analysis of variance (ANOVA), with Holm-Sidak repeated measures multiple comparisons test for the retinal thickness measurements over time. ANOVA with Dunnett's post hoc test was performed for the segmental OCT thickness measurements (SigmaPlot V12, San Jose, California). Differences were considered significant at p < 0.05. For the proteomic studies, the correlation of spectral counts and retinal thickness change at 24 h post IVT injection were analyzed using the Pearson correlation coefficient r.

3. Results

We examined and compared the effects of intravitreal injections of BK, VEGF, and saline control on retinal vessel diameter and tortuosity in rats using SD-OCT. Representative en face retinal images at baseline and at 24 h post injection show retinal vascular dilation and increased tortuosity in eyes subjected to BK and VEGF injections compared with

Fig. 2. Time course of Bradykinin- and VEGF-induced retinal thickening in rats

(A) The effect of intravitreal injection of VEGF, BK and BK/VEGF co-injection on retinal thickness over a time course of 6 h to 5 days using SD-OCT. *p < 0.05, **p < 0.005 compared to baseline using one-way ANOVA with Holm-Sidak repeated measures multiple comparisons test, n = 7 animals per group. (B) Layer thickness measured in each retinal segment of PBS (n = 6), BK (n = 6) and VEGF (n = 6) injected retina using calibrated calipers. The mea-surements were averaged to produce a single thickness value for each retinal layer. NFL: Nerve Fiber Layer, IPL: inner plexiform layer, INL: inner nuclear layer, ONL: outer nuclear layer, SEG: photoreceptor segments, *p < 0.05, **p < 0.005 compared to baseline to 24 h post injection using one-way ANOVA with Dunnett's post hoc test, n = 7 animals per group.

3

fig. 2. Time course of Bradykinin- and VEGF-induced retinal thickening in rats

(a) The effect of intravitreal injection of VEGF, BK and BK/VEGF co-injection on retinal thickness over a time course of 6 h to 5 days using SD-OCT. *p < 0.05, **p < 0.005 compared to baseline using one-way ANOVA with Holm-Sidak repeated measures multiple comparisons test, n = 7 animals per group. (b) Layer thickness measured in each retinal segment of PBS (n = 6), BK (n = 6) and VEGF (n = 6) injected retina using calibrated calipers. The measurements were averaged to produce a single thickness value for each retinal layer. NFL: Nerve Fiber Layer, IPL: inner plexiform layer, INL: inner nuclear layer, ONL: outer nuclear layer, SEG: photoreceptor segments, *p < 0.05, **p < 0.005 compared to baseline to 24 h post injection using one-way ANOVA with Dunnett’s post hoc test, n = 7 animals per group.

Rat retinal samples (n = 4 rats for BK-injected eyes, n = 3 rats for PBS-injected eyes) were harvested 24 h after IVT injection, lysed in TPER tissue extraction reagent (Thermo Scientific, Rockford, IL), and solubilized proteins were separated by SDS-PAGE. Proteins were stained with Coomassie Brilliant Blue (Bio-Rad, Hercules, CA, USA), gel lanes were cut into 40 equal

(7)

Chapter 2.2

slices, and tryptic digests (Sequencing grade modified trypsin, porcine, Promega, Madison, WI) were prepared from each slice, as described previously (Gao et al., 2009). Tryptic digests of each rat retina were analyzed individually by nanospray liquid chromatography tandem mass spectrometry (MS/MS) using a LTQ linear ion trap mass spectrometer (ThermoFisher, Waltham, MA). MS/ MS spectra were analyzed using X!Tandem (The Global Proteome Ma- chine Organization, thegpm.org; version X! Tandem) and the rat se- quence database uni.ReverseRatConcat.2015_03.fasta.pro database. Scaffold 4.4.1.1 (Proteome Software Inc., Portland, OR) was used to analyze and compile MS/MS based peptide and protein identifica-tions. X!Tandem search parameters included the following: maximum valid expectation value of 0.1; residue mass modifications of +16.0 Da for oxidized methionine and +71.0 Da for acrylamide alkylated cysteine; fragment monoisotopic mass error of ± 0.4 Da and precursor mono-isotopic mass error of ± 0.5 Da. Peptide identifications were accepted if they could be established at greater than 90.0% probability with 0.01% false discovery rate (FDR) by the Scaffold Local FDR algorithm. Search results were compiled into a MySQL database and analyzed using MS Results Manager as described previously (Gao et al., 2008). Protein identifications were accepted if they could be established at greater than 99.0% probability and contained at least 2 identified peptides with 0.1% FDR in the same or adjacent gel slices. Protein probabilities were assigned by the Protein Prophet algorithm (Nesvizhskii et al., 2003) and proteins were annotated with Gene Ontology (GO) (Ashburner et al., 2000) terms from NCBI using DAVID Bioinformatics Resources 6.8, NIAID/NIH.

Vitreous samples (n = 4 per group) were obtained from patients during surgery for macular hole (MH) or DME with nonproliferative diabetic retinopathy, as described previously (Kita et al., 2015). Rat retinal extracts (n = 5 per group) were prepared from perfused retina using TPER (Thermo Scientific, Waltham, MA) with protease inhibitor cocktail (Sigma, Milwaukee, WI). Western blotting was performed using rabbit anti-C3 antibody (Abcam, Cambridge, MA). The results were then visualized by enhanced chemiluminescence (20x Lumiglo, Cell Signal-ing, Danvers, MA) and quantified using ImageJ software (NIH, Bethesda, MD).

2.6. Statistics

All results were expressed as mean ± standard error of the mean (SEM). The statistical significance of differences between groups was analyzed by one way analysis of variance (ANOVA), with Holm-Sidak repeated measures multiple comparisons test for the retinal thickness measurements over time. ANOVA with Dunnett’s post hoc test was performed for the segmental OCT thickness measurements (SigmaPlot V12, San Jose, California). Differ-ences were considered significant at p < 0.05. For the proteomic studies, the correlation of spectral counts and retinal thickness change at 24 h post IVT injection were analyzed using the Pearson correlation coefficient r.

(8)

43 C h ap te r 2. 2 3. results

We examined and compared the effects of intravitreal injections of BK, VEGF, and saline control on retinal vessel diameter and tortuosity in rats using SD-OCT. Representative en face retinal images at baseline and at 24 h post injection show retinal vascular dilation and increased tortuosity in eyes subjected to BK and VEGF injections compared with retinal ves-sels in eye receiving PBS control injection (Fig. 1A). Retinal vessel diameters were 1.3- and 1.4-fold greater in BK- and VEGF-injected eyes, respectively, compared to eyes receiving a PBS-injection (PBS: 50 µm ± 0.5 µm versus VEGF: 69 µm ± 2 µm and BK: 63 µm ± 3 µm, p < 0.05; Fig. 1B). Intravitreal injections of BK or VEGF similarly increased retinal vessel tortuosity to 59 µm ± 2 µm and 60 µm ± 3 µm, respectively, compared with 27 µm ± 3 µm in eyes receiving PBS (P < 0.05; Fig. 1C).

Next, we investigated the effect of intravitreal injection of VEGF and BK alone and in com-bination on retinal thickness over a time course of 6 h to 5 days using SD-OCT. BK and VEGF similarly increased retinal thickening by 46 µm (24.6%, p = 0.015) versus 39 µm (20.3%, p = 0.004), respectively at 24 h after intravitreal injection, and retinal thicknesses normalized to baseline at 96 h post-injection (Fig. 2). The effects of co-injection with BK and VEGF on retinal thickness were comparable to the effects of either BK or VEGF injections alone.

retinal vessels in eye receiving PBS control injection (Fig. 1A). Retinal vessel diameters were 1.3- and 1.4-fold greater in BK- and VEGF-in-jected eyes, respectively, compared to eyes receiving a PBS-injection (PBS: 50 μm ± 0.5 μm versus VEGF: 69 μm ± 2 μm and BK: 63 μm ± 3 μm, p < 0.05;Fig. 1B). Intravitreal injections of BK or VEGF similarly increased retinal vessel tortuosity to 59 μm ± 2 μm and 60 μm ± 3 μm, respectively, compared with 27 μm ± 3 μm in eyes receiving PBS (P < 0.05;Fig. 1C).

Next, we investigated the effect of intravitreal injection of VEGF and BK alone and in combination on retinal thickness over a time course of 6 h to 5 days using SD-OCT. BK and VEGF similarly increased retinal thickening by 46 μm (24.6%, p = 0.015) versus 39 μm (20.3%, p = 0.004), respectively at 24 h after intravitreal injection, and retinal thicknesses normalized to baseline at 96 h post-injection (Fig. 2). The effects of co-injection with BK and VEGF on retinal thickness were comparable to the effects of either BK or VEGF injections alone. 3.2. Effects of Bradykinin on the retinal proteome

Using mass spectrometry-based proteomics, we identified a total of 1757 proteins in rat retina, including 1725 and 1739 proteins from PBS-and BK-injected eyes, respectively (Supplementary Table 1). We iden-tified 50 proteins (2.8% of total proteome) that were differentially abundant in the BK group compared with the PBS group (Fig. 3). The spectral counts of 18 proteins were decreased by more than 50% (p < 0.05) and increased by > 2-fold (p < 0.05) for 32 proteins in retina from BK-injected eyes versus PBS injected eyes (Tables 1 and 2). The correlations of abundance changes for each protein inTables 1 and 2to the respective retinal thickness changes from baseline to 24 h for each retina analyzed post IVT injection were determined (Supplemental Fig. 1). This analysis revealed that 28 proteins demonstrated a positive correlation (Pearson correlation coefficient, r > 0.7) between their respective spectral abundance and the change in retina thickness with BK injection measured for each eye. This list included those proteins increased in both BK-stimulated rat retina and DME vitreous (Table 3). Gene Ontology (GO) slim terms were used to annotate proteins that were differentially abundant in retina from BK- and saline-injected eyes (Ashburner et al., 2000). Increased levels of both intracellular (n = 17) and extracellular (n = 15) proteins, annotated by GO terms, were identified in the retina from BK-injected eyes. Fourteen of the thirty-two proteins that increased were plasma proteins, including

hemopexin, haptoglobin, transferrin, complement C3, and serpin A3N. The GO slim term biological process terms distribution for the proteins that were increased in retina in BK-injected eyes included ‘response to hormone’ (37%), ‘negative regulation of endopeptidase activity’ (19%), ‘acute-phase response’ (19%), and ‘triglyceride catabolic process (9.4%).

All proteins (n = 18) that were significantly decreased in BK-in-jected eyes were annotated as intracellular enzymes and matrix teins. The analyses of the GO biological process terms revealed 2 pro-teins associated with bicellular tight junction, including filamin A and actinin alpha 4, that decreased in BK injected eyes.

3.3. Comparison of BK-induced changes in the retinal proteome with the DME vitreous proteome

In a previous study, we identified 167 proteins in human vitreous, including 30 proteins that were increased in DME vitreous by 4-fold or more compared with vitreous samples from macular hole (MH, P < 0.001) (Kita et al., 2015). Using gene symbols to compare pro-teomes, we identified 27 proteins that were present in both the DME vitreous (Kita et al., 2015) and the rat retinal proteome (Supplemental Table 1). Further comparison revealed eight proteins that increased both in the BK-stimulated rat retinal and DME vitreous proteomes by > 2-fold (p < 0.05) compared to control rat retina and MH pro-teomes, respectively. These proteins included afamin, haptoglobin, apolipoprotein A-I, apolipoprotein A-IV, complement component C3, alpha 2-HS glycoprotein, hemopexin and serpin peptidase inhibitor A member 1 (Table 3). Five of these proteins have GO term annotations for inflammatory response.

3.4. Analysis of complement C3 in rat retina and vitreous and DME vitreous The mean spectral counts for C3 in retina from PBS and BK injected eyes were 3.33 ± 2.85 and 23.35 ± 6.28 counts respectively (Fig. 4A). The spectral counts for C3 in MH vitreous was 94 ± 31.79 and DME vitreous was 440.9 ± 37.57 (Fig. 4A). These changes in complement C3 protein levels in rat retina and DME vitreous were further investigated by western blotting. C3 protein in retina and vitr-eous from BK injected eyes increased by 5.2-fold (normalized to β-actin levels; p = 0.12) and 4.3-fold (p = 0.02) compared with PBS injected eyes (n = 5 per group,Fig. 4B). Similarly, western blotting revealed

Fig. 3. Bradykinin induced protein changes in the rat retinal proteome. Schematic volcano plot comparing retinal proteomes from eyes injected with BK and

saline (PBS control). A total of 1757 proteins were identified and quantified using spectral counts. Eighteen proteins were decreased (< 0.5-fold, p < 0.05) and 32 proteins were increased (> 2-fold, p < 0.05) in retina from BK-injected eyes (n = 4) compared with retina from eyes receiving saline (PBS) injection (n = 3). The volcano plot shows the distribution of fold change and p values comparing protein levels in retina from BK- and saline-injected eyes (p < 0.05).

4

fig. 3. Bradykinin induced protein changes in the rat retinal proteome. Schematic volcano plot

compar-ing retinal proteomes from eyes injected with BK and saline (PBS control). A total of 1757 proteins were identified and quantified using spectral counts. Eighteen proteins were decreased (< 0.5-fold, p < 0.05) and 32 proteins were increased (> 2-fold, p < 0.05) in retina from BK-injected eyes (n = 4) compared with retina from eyes receiving saline (PBS) injection (n = 3). The volcano plot shows the distribution of fold change and p values comparing protein levels in retina from BK- and saline-injected eyes (p < 0.05).

3.2. Effects of Bradykinin on the retinal proteome

Using mass spectrometry-based proteomics, we identified a total of 1757 proteins in rat retina, including 1725 and 1739 proteins from PBS- and BK-injected eyes, respectively (Supplementary Table 1). We identified 50 proteins (2.8% of total proteome) that were

(9)

differ-Chapter 2.2

t

ab

le 1.

Pr

oteins increased in retina fr

om bradykinin -injected ey es ccession o. g ene s ymbol p rotein _name n egativ e r egulation of _endope ptidase a ctivity a cute-phase _response t rigly ceride c atabolic _process Inflammator y r esponse APO A4 A polipopr otein A4 * 0.028 E X AFM Afamin * 0.039 E MGLL Mono gl yceride lipase * 0.027 I X MUG1 Murino globulin-1 25.40 0.043 E X X X O XR1 Oxidation resistance 1 24.39 0.021 I SERPIN A3N Serine pr

otease inhibitor A3N

19.39 0.0034 E X X X ALDH3A1 Aldeh yde deh ydr ogenase 3 F amil y Member A1 12.01 0.038 I X Y2 PPM1B Pr otein phosphatase , Mg2+/Mn2+ De pendent 1B 10.72 0.033 I PPP1R11 Pr

otein phosphatase 1 Regulator

y Inhibitor Subunit 11 9.39 0.03 I NFU1 NFU1 Ir

on-Sulfur Cluster Scaf

fold 8.83 0.023 I HPX Hemopexin 7.28 0.012 E C3 Complement C3 7.05 0.027 E X X X MGEA5 Pr otein O-GlcN Acase 6.90 0.017 I X LOC299282 Ser pin CS 6.46 0.0007 E AHSG Alpha-2-HS-gl ycopr otein 5.07 0.038 E X X X X Ppp2r5e Pr

otein Phosphatase 2 Regulator

y Subunit B’ Epsilon 4.78 0.043 I APO A1 A polipopr otein A1 4.46 0.035 E X PPM1A Pr otein phosphatase , Mg2+/Mn2+ De pendent 1A 4.26 0.037 I HP Hapto globin 4.15 0.0052 E X X X

(10)

45 C h ap te r 2. 2 t ab le 1. Pr

oteins increased in retina fr

om bradykinin -injected ey es (continued) a ccession _no. g ene s ymbol p rotein _name n egativ e r egulation of _endope ptidase a ctivity a cute-phase _response t rigly ceride c atabolic _process Inflammator y r esponse P17475 SERPIN A1 Ser pin F amil y A Member 1 4.10 0.05 E X X X X D3ZB78 R GD1559864 PHD finger pr otein 24 3.85 0.0082 I Q9QZ76 MB My oglobin 3.77 0.022 I X P0C1X8 AAK1 AP2-associated kinase 1 3.65 0.0072 I P68101 Eif2s1 Eukar

yotic translation initiation factor 2

subunit α 3.31 0.0049 I P48199 CRP C-reacti ve pr otein 3.30 0.0057 E X X X Q6IRS6 FETUB Fetuin B 3.11 0.033 E X M0RA39 MXRA7

Matrix remodeling associated 7

3.06 0.013 E R9PXR7 PT GES3 Pr ostaglandin E synthase 3 2.71 0.0084 I X G3V6H9 N AP1L1 Nucleosome assembl y pr otein 1-lik e 1 2.65 0.032 I X Q62876 SYNGR1 Synapto gyrin-1 2.33 0.036 I F1LQZ0 TMEM65 T ransmembrane pr otein 65 2.24 0.024 I Q7TMC7 TF T ransfer rin 2.24 0.0017 E X X X (*denotes pr oteins that w ere not detected rat retina fr om ey es injected with PBS , p values deter mined using t-test). BK-injected ey es (n=4), contr ol PBS injected ey es (n=3).

(11)

Chapter 2.2

entially abundant in the BK group compared with the PBS group (Fig. 3). The spectral counts of 18 proteins were decreased by more than 50% (p < 0.05) and increased by > 2-fold (p < 0.05) for 32 proteins in retina from BK-injected eyes versus PBS injected eyes (Tables 1 and 2). The correlations of abundance changes for each protein in Tables 1 and 2 to the respective retinal thickness changes from baseline to 24 h for each retina analyzed post IVT injection were determined (Supplemental Fig. 1). This analysis revealed that 28 proteins demonstrated a positive correlation (Pearson correlation coefficient, r > 0.7) between their respective spectral abundance and the change in retina thickness with BK injection measured for each eye. This list included those proteins increased in both BK-stimulated rat retina and DME vitreous (Table 3). Gene Ontology (GO) slim terms were used to annotate proteins that were differentially abundant in retina from BK- and saline-injected eyes (Ashburner et al., 2000). Increased levels of both intracellular (n = 17) and extracellular (n = 15) proteins, annotated by GO terms, were identified in the retina from BK-injected eyes. Fourteen of the thirty-two proteins that increased were plasma proteins, including hemopexin, haptoglobin, transferrin,

table 2. Proteins decreased in retina from bradykinin -injected eyes accession no. gene symbol protein name fold change (bk/pbs) p-value cytoskeletal enzymatic activity D3ZRN5 TROVE2 TROVE domain family member 2 0.49 0.026

G3V6L9 FKBP3 FK506 binding protein 3 0.46 0.039 X

B0K031 RPL7 Ribosomal protein L7 0.43 0.017

Q9QXQ0 ACTN4 Actinin alpha 4 0.41 0.044 X

M0R517 LRRC74B Leucine-rich repeat-containing 74B 0.37 0.021

P62859 RPS28 Ribosomal protein S28 0.35 0.007

C0JPT7 FLNA Filamin A 0.33 0.037 X

Q812D3 PPIL3 Peptidyl-prolyl isomerase-like 3 0.28 0.025 X

P29457 SERPINH1 Serpin family h member 1 0.26 0.036

Q641Z6 EHD1 EH domain-containing 1 0.26 0.0092

Q62991 SCFD1 Sec1 family domain-containing 1 0.23 0.039

D3ZGX7 MTDH Metadherin 0.20 0.047

F1M6T7 ERICH5 Glutamate rich 5 0.20 0.016

F1LW91 NUMA1 Nuclear mitotic apparatus protein 1 0.18 0.032 X Q5M9H2 ACADVL Acyl-Coenzyme A dehydrogenase,

very long chain

0.15 0.029 X

Q4V8K5 BROX BRO1 domain and CAAX motif containing

0.13 0.029

Q6AYK8 EIF3D Eukaryotic translation initiation factor 3 subunit D

* 0.025

Q5U2R7 MESDC2 Mesoderm development candidate 2 * 0.0002

(*denotes proteins that were not detected rat retina from eyes injected with PBS, p values determined using t-test). BK-injected eyes (n=4), control PBS injected eyes (n=3).

(12)

47 C h ap te r 2. 2

complement C3, and serpin A3N. The GO slim term biological process terms distribution for the proteins that were increased in retina in BK-injected eyes included ‘response to hormone’ (37%), ‘negative regulation of endopeptidase activity’ (19%), ‘acute-phase response’ (19%), and ‘triglyceride catabolic process (9.4%).

All proteins (n = 18) that were significantly decreased in BK-injected eyes were annotated as intracellular enzymes and matrix proteins. The analyses of the GO biological process terms revealed 2 proteins associated with bicellular tight junction, including filamin A and actinin alpha 4, that decreased in BK injected eyes.

3.3. Comparison of BK-induced changes in the retinal proteome with the DME vitreous proteome

In a previous study, we identified 167 proteins in human vitreous, including 30 proteins that were increased in DME vitreous by 4-fold or more compared with vitreous samples from macular hole (MH, P < 0.001) (Kita et al., 2015). Using gene symbols to compare proteomes, we identified 27 proteins that were present in both the DME vitreous (Kita et al., 2015) and the rat retinal proteome (Supplemental Table 1). Further comparison revealed eight proteins that increased both in the BK-stimulated rat retinal and DME vitreous proteomes by > 2-fold (p < 0.05) compared to control rat retina and MH proteomes, respectively. These proteins included afamin, haptoglobin, apolipoprotein A-I, apolipoprotein A-IV, complement component C3, alpha 2-HS glycoprotein, hemopexin and serpin peptidase inhibitor A member 1 (Table 3). Five of these proteins have GO term annotations for inflammatory response.

table 3. Proteins that are increased in both retina from rats with bradykinin-induced retinal edema and

DME vitreous. The table lists the fold change in spectral counts for proteins that increased (> 2-fold, p < 0.05) in both rat retina from BK injected eyes (n=4) compared with PBS injected eyes (n=3) and hu-man vitreous samples from subjects with DME compared subjects with macular hole (MH). (* denotes proteins that were not detected rat retina from eyes injected with PBS, p values determined using t-test)

gene symbol protein name fold change bk/pbs (rat retina) p-value bk/pbs (rat retina) fold change dme/mh (human vitreous) p value dme/mh (human vitreous)

AFM Afamin * 0.039 7.1 2.55E-06

HP Haptoglobin 4.2 0.0052 6 6.12E-06

APOA1 Apolipoprotein A1 4.5 0.035 5.7 3.65E-07

APOA4 Apolipoprotein A4 * 0.028 4.8 6.77E-05

C3 Complement C3 7.1 0.027 4.7 1.34E-05

AHSG Alpha 2-HS-glycoprotein 5.1 0.038 4.2 3.88E-04

HPX Hemopexin 7.3 0.012 2.7 8.35E-07

(13)

Chapter 2.2

48

3.4. Analysis of complement C3 in rat retina and vitreous and DME vitreous

The mean spectral counts for C3 in retina from PBS and BK injected eyes were 3.33 ± 2.85 and 23.35 ± 6.28 counts respectively (Fig. 4A). The spectral counts for C3 in MH vitreous was 94 ± 31.79 and DME vitreous was 440.9 ± 37.57 (Fig. 4A). These changes in complement C3 protein levels in rat retina and DME vitreous were further investigated by western blotting. C3 protein in retina and vitreous from BK injected eyes increased by 5.2-fold (normalized to β-actin levels; p = 0.12) and 4.3-fold (p = 0.02) compared with PBS injected eyes (n = 5 per group, Fig. 4B). Similarly, western blotting revealed that C3 protein in the DME vitreous was increased by 6.6-fold (p = 0.039) compared with MH vitreous samples (n = 4 per group, Fig. 4C).

4. dIscussIon

This study utilized an experimental rat model of BK-induced retinal thickening to identify proteomic changes in the retina that are associated with retinal edema. Of the 1757 proteins detected in the rat retina, we identified 50 proteins that differ in abundance in retina with BK-induced retinal edema compared to saline-injected controls. To our knowledge, this is the first description of a retinal proteome in an experimental model of retinal edema.

In this longitudinal study, we show that a single IVT injection of BK or VEGF similarly and transiently increase retinal thickness. Both BK and VEGF increase IPL and ONL thickness

induced retinal edema were expected. Of note, we did not observe a significant increase in albumin, which is often used as a marker for retinal vascular leakage (Kielczewski et al., 2011;Vinores et al., 1989). All proteins that significantly decreased in BK-injected eyes were annotated as intracellular proteins, including three cytoskeletal pro-teins including actinin alpha 4 (Patrie et al., 2002) and filamin A (Feng et al., 2006). These proteins interact with tight junction and adherens-type junction components and thereby may play a role blood retinal barrier function, matrix remodeling, and macular edema (Cehofski et al., 2015). These results suggest that BK's effects on the retina might be mediated, in part, by changes in cytoskeletal proteins.

Comparison of the retinal proteome from rats with BK-induced retinal edema with the vitreous proteome from DME (Kita et al., 2015) revealed eight plasma proteins that were increased in both conditions. Five proteins identified in both rat retina with BK-induced edema and vitreous samples from DME are annotated with inflammatory functions (haptoglobin, apolipoprotein A-IV, complement C3, alpha-2-HS-glyco-protein, alpha-1-antiproteinase Serpina1). In contrast, as mentioned above, only 2 proteins (AAK1 and EiF2s1) were increased both in retina from rats with BK-induced retinal edema and in retina from diabetic mice (Gao et al., 2009). Further studies using models of retinal edema in rodents may provide opportunities in addition to diabetes alone to evaluate protein changes in the retina that may contribute to DME.

Western blotting confirmed that C3 levels were increased in retina and vitreous from BK injected eyes and in DME vitreous. C3 is an abundant plasma protein that is also expressed in retinal pigment epi-thelial cells (Luo et al., 2013). Complement C3 is an essential protein in the activation of the complement system which is a key component of innate immunity (Sahu and Lambris, 2001). Our findings are consistent with previous reports demonstrating that complement C3 is increased in the vitreous of persons with advanced stages of diabetic retinopathy (Gao et al., 2008;Garcia-Ramirez et al., 2007). Physiological levels of

pathophysiological increases in C3 has been implicated in retinal de-generation (Bosco et al., 2018;Cashman et al., 2011;Natoli et al., 2017) and neuroinflammation (Alawieh et al., 2018;Litvinchuk et al., 2018). Since C3 levels are increased in DME, further studies are warranted to characterize the potential effect(s) of C3 in retinal edema and neuror-etinal function.

The proteomic approach used here has significant limitations. First, the detection sensitivity of the mass spectrometry-based proteomics method used in this study is less than other methods, such as im-munodetection, and mainly evaluates relatively abundant proteins. Second, spectral counts were used as a measure of protein abundance for both human vitreous and rat retinal samples. This method is semi-quantitative and is influenced by experimental variability in peptide recovery and detection. Post-translational modifications of peptides that are not specified in the analytic methods, described above, will reduce spectral count. Third, retinal proteomes were analyzed from a relatively small number of samples. To reduce type 1 errors [false po-sitives] that influence the discovery, we used peptide and protein thresholds of FDR of 0.01% and 0.1%, respectively and an additional threshold of > 2-fold increase or more than 50% decrease in the BK group compared with control. Finally, gene symbols were used to compare proteins across 3 proteomic studies involving markedly dif-ferent conditions, namely rat retina with BK-induced edema, retina from mice with diabetes, and human DME vitreous (Gao et al., 2009;

Kita et al., 2015). Although these three proteomic studies were per-formed using the same mass spectrometer and spectral counting methods, differences in tissue processing and amino acid sequences across species could affect protein identification and quantitation. Therefore, comparisons of proteomes across species should be con-sidered with these limitations.

In summary, this study characterized the rat retinal ultrastructure and protein composition in an acute model of retinal thickening

in-Fig. 4. Proteomic and Western blot analyses of Complement C3 protein in retina and vitreous. (A) C3 spectral counts in human DME and MH vitreous samples

and in rat retina 24 h after PBS (n = 3) and BK (n = 4) injection (B) Western blotting for C3 protein levels in the rat vitreous and rat retina 24 h after BK and PBS injections (n = 5). (C) Western blotting for C3 protein levels in vitreous samples from DME and MH (n = 4). P values determined using t-test.

fig. 4. Proteomic and Western blot analyses of Complement C3 protein in retina and vitreous. (a) C3

spectral counts in human DME and MH vitreous samples and in rat retina 24 h after PBS (n = 3) and BK (n = 4) injection (b) Western blotting for C3 protein levels in the rat vitreous and rat retina 24 h after BK and PBS injections (n = 5). (c) Western blotting for C3 protein levels in vitreous samples from DME and MH (n = 4). P values determined using t-test.

(14)

49 C h ap te r 2. 2

along with a trend for thickening in the INL without the appearance of cystoid-like lesions. This rodent model of retinal edema differs from macular edema in humans where intra- and sub-retinal cysts are frequently observed. Moreover, the extent of retinal thickening is markedly less than that observed in DME. The short and transient nature of BK-induced retinal edema in rats differs from the chronic condition of macular edema in humans and is a limitation of this experimental animal model. In addition, we show that BK and VEGF also similarly increase retinal vessel diameters and tortuosity. Increased retinal vessel diameters and tortuosity in diabetes and DME have been previously reported (Drobnjak et al., 2017; Kristinsson et al., 1997; Sasongko et al., 2011).

We identified 32 proteins from a total of 1757 proteins that were increased by > 2-fold in rat retina with BK-induced edema. In a previous study, 65 proteins from a total of 1792 proteins identified were increased by > 2-fold in retina from mice with diabetes compared with nondiabetic controls (Gao et al., 2009). Comparison of the retinal proteome changes associated with BK-induced edema in rats and diabetes in mice revealed only 2 proteins, namely AP2-associated protein kinase 1 and eukaryotic translation initiation factor 2 subunit 1, that were increased in both studies. Moreover, nearly all the proteins identified as increased in diabetic retina proteome were intracellular or membrane associated (Gao et al., 2009) whereas 15 of the 32 proteins (47%) identified as increase in BK-induced retinal edema are extracellular. None of the BK-induced changes in retinal proteins were observed in the streptozotocin-induced diabetes associated changes in a study of 527 proteins in the retinal proteome (VanGuilder et al., 2011). Matrix-remodeling associated protein 7 was identified as increased in BK-induced retinal edema in this study (Table 1) was also observed to be increased in the retinal proteome of diabetic db/db mice compared with nondiabetic controls (Ly et al., 2014). With few exceptions, there were marked differences in the protein changes detected in retinal proteomes in acute BK-induced edema and in long-term diabetes in ro-dents. This could be associated with difference in retinal ultrastructure of these two models where BK causes retinal thickening whereas diabetes causes retinal thinning (Kita et al., 2015). Moreover, since BK increases RVP, elevated levels of plasma proteins in BK-induced retinal edema were expected. Of note, we did not observe a significant increase in albumin, which is often used as a marker for retinal vascular leakage (Kielczewski et al., 2011; Vinores et al., 1989). All proteins that significantly decreased in BK-injected eyes were annotated as intracellular proteins, including three cytoskeletal proteins including actinin alpha 4 (Patrie et al., 2002) and filamin A (Feng et al., 2006). These proteins interact with tight junction and adherens- type junction components and thereby may play a role blood retinal barrier function, matrix remodeling, and macular edema (Cehofski et al., 2015). These results suggest that BK’s effects on the retina might be mediated, in part, by changes in cytoskeletal proteins.

Comparison of the retinal proteome from rats with BK-induced retinal edema with the vitreous proteome from DME (Kita et al., 2015) revealed eight plasma proteins that were in-creased in both conditions. Five proteins identified in both rat retina with BK-induced edema and vitreous samples from DME are annotated with inflammatory functions (haptoglobin,

(15)

Chapter 2.2

apolipoprotein A-IV, complement C3, alpha-2-HS-glycoprotein, alpha-1-antiproteinase Ser-pina1). In contrast, as mentioned above, only 2 proteins (AAK1 and EiF2s1) were increased both in retina from rats with BK-induced retinal edema and in retina from diabetic mice (Gao et al., 2009). Further studies using models of retinal edema in rodents may provide opportunities in addition to diabetes alone to evaluate protein changes in the retina that may contribute to DME.

Western blotting confirmed that C3 levels were increased in retina and vitreous from BK injected eyes and in DME vitreous. C3 is an abundant plasma protein that is also expressed in retinal pigment epithelial cells (Luo et al., 2013). Complement C3 is an essential protein in the activation of the complement system which is a key component of innate immunity (Sahu and Lambris, 2001). Our findings are consistent with previous reports demonstrating that complement C3 is increased in the vitreous of persons with advanced stages of diabetic retinopathy (Gao et al., 2008; Garcia-Ramirez et al., 2007). Physiological levels of comple-ment C3 has protective effects on photoreceptor loss and Bruch’s membrane thickening during aging (Hoh Kam et al., 2013) whereas pathophysiological increases in C3 has been implicated in retinal degeneration (Bosco et al., 2018; Cashman et al., 2011; Natoli et al., 2017) and neuroinflammation (Alawieh et al., 2018; Litvinchuk et al., 2018). Since C3 levels are increased in DME, further studies are warranted to characterize the potential effect(s) of C3 in retinal edema and neuroretinal function.

The proteomic approach used here has significant limitations. First, the detection sensitiv-ity of the mass spectrometry-based proteomics method used in this study is less than other methods, such as immunodetection, and mainly evaluates relatively abundant proteins. Sec-ond, spectral counts were used as a measure of protein abundance for both human vitreous and rat retinal samples. This method is semi-quantitative and is influenced by experimental variability in peptide recovery and detection. Post-translational modifications of peptides that are not specified in the analytic methods, described above, will reduce spectral count. Third, retinal proteomes were analyzed from a relatively small number of samples. To reduce type 1 errors [false positives] that influence the discovery, we used peptide and protein thresholds of FDR of 0.01% and 0.1%, respectively and an additional threshold of > 2-fold increase or more than 50% decrease in the BK group compared with control. Finally, gene symbols were used to compare proteins across 3 proteomic studies involving markedly different conditions, namely rat retina with BK-induced edema, retina from mice with diabetes, and human DME vitreous (Gao et al., 2009; Kita et al., 2015). Although these three proteomic studies were performed using the same mass spectrometer and spectral counting methods, differences in tissue processing and amino acid sequences across species could affect protein identification and quantitation. Therefore, comparisons of proteomes across species should be considered with these limitations.

In summary, this study characterized the rat retinal ultrastructure and protein composition in an acute model of retinal thickening induced by intravitreal injection of BK. Mass spec-trometry based proteomic analysis of the rat retina identified pro-inflammatory molecules

(16)

51 C h ap te r 2. 2

and plasma proteins that were elevated in the BK injected rat eyes, including 8 proteins that were previously reported to be increased in human DME vitreous proteome. These results provide insight into the proteomic changes in the retina that are associated with retinal edema.

(17)

Chapter 2.2 references

Abdouh, M., Khanjari, A., Abdelazziz, N., Ongali, B., Couture, R., Hassessian, H.M., 2003. Early up-regulation of kinin B1 receptors in retinal microvessels of the streptozotocin- diabetic rat. Br. J. Pharmacol. 140, 33–40.

Alawieh, A., Langley, E.F., Weber, S., Adkins, D., Tomlinson, S., 2018. Identifying the role of complement in triggering neuroinflammation after traumatic brain injury. J. Neurosci. 38, 2519–2532. Ashburner, M., Ball, C.A., Blake, J.A., Botstein, D., Butler, H., Cherry, J.M., Davis, A.P.,

Dolinski, K., Dwight, S.S., Eppig, J.T., Harris, M.A., Hill, D.P., Issel-Tarver, L., Kasarskis, A., Lewis, S., Matese, J.C., Richardson, J.E., Ringwald, M., Rubin, G.M.,

Sherlock, G., 2000. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat. Genet. 25, 25–29.

Bosco, A., Anderson, S.R., Breen, K.T., Romero, C.O., Steele, M.R., Chiodo, V.A., Boye, S.L., Hauswirth, W.W., Tomlinson, S., Vetter, M.L., 2018. Complement C3-targeted gene therapy restricts onset and progression of neurodegeneration in chronic mouse

glaucoma. Mol. Ther. 26, 2379–2396.

Cashman, S.M., Desai, A., Ramo, K., Kumar-Singh, R., 2011. Expression of complement component 3 (C3) from an adenovirus leads to pathology in the murine retina.

Investig. Ophthalmol. Vis. Sci. 52, 3436–3445.

Cehofski, L.J., Kruse, A., Kjærgaard, B., Stensballe, A., Honoré, B., Vorum, H., 2015.

Proteins involved in focal adhesion signaling pathways are differentially regulated in experimental branch retinal vein occlusion. Exp. Eye Res. 138, 87–95.

Clermont, A., Chilcote, T.J., Kita, T., Liu, J., Riva, P., Sinha, S., Feener, E.P., 2011. Plasma kallikrein me-diates retinal vascular dysfunction and induces retinal thickening in diabetic rats. Diabetes 60, 1590–1598.

Clermont, A., Murugesan, N., Zhou, Q., Kita, T., Robson, P.A., Rushbrooke, L.J., Evans, D.M., Aiello, L.P., Feener, E.P., 2016. Plasma kallikrein mediates vascular endothelial

growth factor-induced retinal dysfunction and thickening. Investig. Ophthalmol. Vis. Sci. 57, 2390–2399. Drobnjak, D., Munch, I.C., Glumer, C., Faerch, K., Kessel, L., Larsen, M., Veiby, N., 2017.

Relationship between retinal vessel diameters and retinopathy in the Inter99 Eye Study. J. Clin. Transl. Endocrinol. 8, 22–28.

Duh, E.J., Sun, J.K., Stitt, A.W., 2017. Diabetic retinopathy: current understanding, mechanisms, and treatment strategies. JCI Insight 2, 1–13.

Feng, Y., Chen, M.H., Moskowitz, I.P., Mendonza, A.M., Vidali, L., Nakamura, F., Kwiatkowski, D.J., Walsh, C.A., 2006. Filamin A (FLNA) is required for cell-cell contact in vascular development and cardiac morphogenesis. Proc. Natl. Acad. Sci. U.S.A. 103, 19836–19841.

Gao, B.B., Chen, X., Timothy, N., Aiello, L.P., Feener, E.P., 2008. Characterization of the

vitreous proteome in diabetes without diabetic retinopathy and diabetes with proliferative diabetic retinopathy. J. Proteome Res. 7, 2516–2525.

Gao, B.B., Clermont, A., Rook, S., Fonda, S.J., Srinivasan, V.J., Wojtkowski, M., Fujimoto,

J.G., Avery, R.L., Arrigg, P.G., Bursell, S.E., Aiello, L.P., Feener, E.P., 2007. Extracellular carbonic anhydrase mediates hemorrhagic retinal and cerebral vascular permeability through prekallikrein activation. Nat. Med. 13, 181–188.

Gao, B.B., Phipps, J.A., Bursell, D., Clermont, A.C., Feener, E.P., 2009. Angiotensin AT1

receptor antagonism ameliorates murine retinal proteome changes induced by diabetes. J. Proteome Res. 8, 5541–5549.

(18)

53 C h ap te r 2. 2

Garcia-Ramirez, M., Canals, F., Hernandez, C., Colome, N., Ferrer, C., Carrasco, E.,

Garcia-Arumi, J., Simo, R., 2007. Proteomic analysis of human vitreous fluid by fluorescence-based dif-ference gel electrophoresis (DIGE): a new strategy for identifying potential candidates in the pathogenesis of proliferative diabetic retinopathy. Diabetologia 50, 1294–1303.

Hoh Kam, J., Lenassi, E., Malik, T.H., Pickering, M.C., Jeffery, G., 2013. Complement

component C3 plays a critical role in protecting the aging retina in a murine model of age-related macular degeneration. Am. J. Pathol. 183, 480–492.

Kielczewski, J.L., Li Calzi, S., Shaw, L.C., Cai, J., Qi, X., Ruan, Q., Wu, L., Liu, L., Hu, P.,

Chan-Ling, T., Mames, R.N., Firth, S., Baxter, R.C., Turowski, P., Busik, J.V., Boulton, M.E., Grant, M.B., 2011. Free insulin-like growth factor binding protein-3 (IGFBP-3)

reduces retinal vascular permeability in association with a reduction of acid sphin- gomyelinase (AS-Mase). Investig. Ophthalmol. Vis. Sci. 52, 8278–8286.

Kita, T., Clermont, A.C., Murugesan, N., Zhou, Q., Fujisawa, K., Ishibashi, T., Aiello, L.P., Feener, E.P., 2015. Plasma kallikrein-kinin system as a VEGF-independent mediator of diabetic macular edema. Diabetes 64, 3588–3599.

Kristinsson, J.K., Gottfredsdottir, M.S., Stefansson, E., 1997. Retinal vessel dilatation and elongation precedes diabetic macular oedema. Br. J. Ophthalmol. 81, 274–278.

Lee, R., Wong, T.Y., Sabanayagam, C., 2015. Epidemiology of diabetic retinopathy, diabetic macular edema and related vision loss. Eye Vis. (Lond). 2, 17–42.

Litvinchuk, A., Wan, Y.W., Swartzlander, D.B., Chen, F., Cole, A., Propson, N.E., Wang, Q., Zhang, B., Liu, Z., Zheng, H., 2018. Complement C3aR inactivation attenuates tau pathology and reverses an im-mune network deregulated in tauopathy models and alzheimer’s disease. Neuron 100, 1337–1353. Liu, J., Clermont, A.C., Gao, B.B., Feener, E.P., 2013. Intraocular hemorrhage causes

retinal vascular dysfunction via plasma kallikrein. Investig. Ophthalmol. Vis. Sci. 54, 1086–1094. Liu, J., Feener, E.P., 2013. Plasma kallikrein-kinin system and diabetic retinopathy. Biol. Chem. 394, 319–328.

Loukovaara, S., Nurkkala, H., Tamene, F., Gucciardo, E., Liu, X., Repo, P., Lehti, K., Varjosalo, M., 2015. Quantitative proteomics analysis of vitreous humor from diabetic retinopathy patients. J. Pro-teome Res. 14, 5131–5143.

Luo, C., Zhao, J., Madden, A., Chen, M., Xu, H., 2013. Complement expression in retinal pigment epithe-lial cells is modulated by activated macrophages. Exp. Eye Res. 112, 93–101.

Ly, A., Scheerer, M.F., Zukunft, S., Muschet, C., Merl, J., Adamski, J., de Angelis, M.H., Neschen, S., Hauck, S.M., Ueffing, M., 2014. Retinal proteome alterations in a mouse model of type 2 diabetes. Diabetologia 57, 192–203.

Ma, J.X., Song, Q., Hatcher, H.C., Crouch, R.K., Chao, L., Chao, J., 1996. Expression and

cellular localization of the kallikrein-kinin system in human ocular tissues. Exp. Eye Res. 63, 19–26. Natoli, R., Fernando, N., Jiao, H., Racic, T., Madigan, M., Barnett, N.L., Chu-Tan, J.A.,

Valter, K., Provis, J., Rutar, M., 2017. Retinal macrophages synthesize C3 and acti- vate complement in AMD and in models of focal retinal degeneration. Investig.

Ophthalmol. Vis. Sci. 58, 2977–2990.

Nesvizhskii, A.I., Keller, A., Kolker, E., Aebersold, R., 2003. A statistical model for identifying proteins by tandem mass spectrometry. Anal. Chem. 75, 4646–4658.

Patrie, K.M., Drescher, A.J., Welihinda, A., Mundel, P., Margolis, B., 2002. Interaction of two actin-binding proteins, synaptopodin and alpha-actinin-4, with the tight junction protein MAGI-1. J. Biol. Chem. 277, 30183–30190.

(19)

Chapter 2.2

Plasma kallikrein mediates angiotensin II type 1 receptor-stimulated retinal vascular permeability. Hy-pertension 53, 175–181.

Riedl, M., 2012. Hereditary angioedema therapies in the United States: movement toward an international treatment consensus. Clin. Ther. 34, 623–630.

Sahu, A., Lambris, J.D., 2001. Structure and biology of complement protein C3, a con- necting link between innate and acquired immunity. Immunol. Rev. 180, 35–48.

Sasongko, M.B., Wong, T.Y., Nguyen, T.T., Cheung, C.Y., Shaw, J.E., Wang, J.J., 2011.

Retinal vascular tortuosity in persons with diabetes and diabetic retinopathy. Diabetologia 54, 2409–2416. VanGuilder, H.D., Bixler, G.V., Kutzler, L., Brucklacher, R.M., Bronson, S.K., Kimball, S.R., Freeman, W.M.,

2011. Multi-modal proteomic analysis of retinal protein expression alterations in a rat model of diabetic retinopathy. PLoS One 6, e16271.

Vinores, S.A., Gadegbeku, C., Campochiaro, P.A., Green, W.R., 1989.

Immunohistochemical localization of blood-retinal barrier breakdown in human diabetics. Am. J. Pathol. 134, 231–235.