Metabolomic profiling of aqueous humor in glaucoma points to taurine

Subscriber access provided by WEBSTER UNIV

Article

Metabolomic profiling of aqueous humor in glaucoma points to taurine and spermine deficiency: findings from the Eye-D study Adrien Buisset, Philippe Gohier, Stéphanie Leruez, Jeanne Muller, Patrizia AmatiBonneau, Guy Lenaers, Dominique Bonneau, Gilles Simard, Vincent Procaccio, Cedric Annweiler, Dan Milea, Pascal Reynie, and Juan Manuel Chao de la Barca J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.8b00915 • Publication Date (Web): 31 Jan 2019 Downloaded from http://pubs.acs.org on February 5, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Metabolomic profiling of aqueous humor in glaucoma points to taurine and spermine deficiency: findings from the Eye-D study Adrien Buisset,† Philippe Gohier,† Stéphanie Leruez,†,‡ Jeanne Muller,† Patrizia AmatiBonneau,‡,‖ Guy Lenaers,‡ Dominique Bonneau,‡,║ Gilles Simard,║ Vincent Procaccio,‡,║ Cédric Annweiler,#,§ Dan Milea,◊ Pascal Reynier, ‡,║,* Juan Manuel Chao de la Barca‡,║ †Département ‡Unité

d’Ophtalmologie, Centre Hospitalier Universitaire, Angers, France

Mixte de Recherche MITOVASC, équipe Mitolab, Centre National de la Recherche

Scientifique 6015, Institut National de la Santé et de la Recherche Médicale U1083, Université d'Angers, Angers, France ║Département #Department

de Biochimie et Génétique, Centre Hospitalier Universitaire, Angers, France

of Geriatric Medicine, Angers University Hospital; Angers University Memory Clinic;

Research Center on Autonomy and Longevity; UPRES EA 4638, University of Angers, France §Robarts

Research Institute, Department of Medical Biophysics, Schulich School of Medicine and

Dentistry, the University of Western Ontario, London, ON, Canada ◊Singapore

Eye Research Institute, Singapore National Eye Centre, Duke-NUS, Singapore

*Corresponding author: Pascal Reynier, Département de Biochimie et Génétique, Centre Hospitalier Universitaire, Angers, France. Email: [email protected] ABBREVIATIONS: AUROC: area under the receiver operating characteristic curve; BMI: body mass index; CCT: central corneal thickness; FIA: flow injection analysis; HPLC: highperformance liquid chromatography; LC-MS: liquid chromatography-mass spectrometry; QC: quality control; LC: liquid chromatography; LLOQ: lower limit of quantitation; LR: logistic regression; OCT: optical coherence tomography; PACG: primary angle-closure glaucoma; PC: phosphatidylcholines; PCA: principal component analysis; PLS-DA: partial least squaresdiscriminant analysis; POAG: primary open-angle glaucoma; RNFL: retinal nerve fibre layer; SM: sphingomyelins; VIP: variables important for the projection; ULOQ: upper limit of quantitation.

1

ACS Paragon Plus Environment

Journal of Proteome Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

ABSTRACT We compared the metabolomic profile of aqueous humor from patients with primary openangle glaucoma (POAG; n = 26) with that of a group of age- and sex-matched non-POAG controls (n = 26), all participants undergoing cataract surgery. Supervised paired partial least squares discriminant analysis showed good predictive performance for test sets with a median area under the receiver operating characteristic of 0.89 and a p-value of 0.0087. Twenty-three metabolites allowed discrimination between the two groups. Univariate analysis after the Benjamini-Hochberg correction showed significant differences for 13 of these metabolites. The POAG metabolomic signature indicated reduced concentrations of taurine and spermine, and increased concentrations of creatinine, carnitine, three short-chain acylcarnitines, 7 amino acids (glutamine, glycine, alanine, leucine, isoleucine, hydroxyl-proline, and acetyl-ornithine), 7 phosphatidylcholines, one lysophosphatidylcholine, and one sphingomyelin. This suggests an alteration of metabolites involved in osmoprotection (taurine and creatinine), neuroprotection (spermine, taurine, and carnitine), amino acid metabolism (7 amino acids and three acylcarnitines), and the remodeling of cell membranes drained by the aqueous humor (hydroxyproline and phospholipids). Four of these metabolic alterations, already reported in POAG plasma, concern spermine, C3 and C4 acylcarnitines, PC aa 34:2 and PC aa 36:4, thus highlighting their importance in the pathogenesis of glaucoma.

KEYWORDS: Aqueous humor, Glaucoma, Lipidomics, Metabolomics, Primary open-angle glaucoma

2


Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION Aqueous humor, the transparent fluid found in the anterior chamber of the eye, has a volume of about 100-200 microliters in human adults. It is secreted by the ciliary epithelium in the posterior chamber before flowing through the pupil into the anterior chamber. Aqueous humor provides nutrients to the surrounding avascular cornea and lens while draining metabolic waste from the eye to the venous blood. The homeostasis of aqueous humor also allows the maintenance of adequate intraocular pressure as well as optimal optical and refractive properties.

The aqueous humor contains a complex mixture of electrolytes and metabolites. Recent studies using liquid chromatography-mass spectrometry (LC-MS) have identified almost 250 metabolites belonging to 47 metabolic pathways, including carbohydrates, amino acids, urea, ascorbic acid, glutathione organic solutes, adenosine, xanthine, polyamines, fatty acids, acylcarnitines, and phospholipids.1,2 An increasing number of metabolomic studies, performed on animal models under pathophysiological conditions, have been aimed at identifying the most important metabolites and lipids involved in various ophthalmic conditions. For example, discriminant aqueous humor metabolomic signatures were found in rabbits after UV-A and UV-B irradiation,3 and in rats after uveitis induced by exposure to endotoxins.4 In humans, diabetes,2 high degree myopia,5,6 and the progression of cataract and keratoconus,7 were also associated with discriminant metabolomic signatures in the aqueous humor.

With regard to glaucoma, the aqueous humor of a rat model with the disease induced by the intracamerular injection of sodium hyaluronate showed a significant metabolomic signature containing 16 discriminant metabolites including lipoproteins, cholesterol, amino acids, fatty 3



Page 4 of 30

acids, glucose and creatine or creatinine.8 A study of the oxygen tension and metabolomic effects of glaucoma drainage devices on the aqueous humor of the rabbit showed the significant impact of citric acid, glucose and tyrosine.9 Phospholipidomics, performed on the DBA/2J and DBA/2J-Gpnmb+ mouse models revealed phospholipidic profiles that varied according to the elevation of the intraocular pressure.10 In humans, only a few lipidomic studies have so far reported the phospholipidic profile of the aqueous humor in patients with primary open-angle glaucoma (POAG) compared to controls,11-13 but to date the metabolomics profile of polar metabolites and other lipids, such as acylcarnitines, has never been reported in glaucoma. Here we present a targeted metabolomic and lipidomic study comparing the aqueous humor of patients with POAG and that of controls, all undergoing cataract surgery.

MATERIAL AND METHODS Study participants We studied 52 participants from the Eye-D study, a French observational case-control study designed to compare the composition of POAG cases’ aqueous humor with that of age- and sex-matched non-POAG controls, the aqueous humor sampling being performed in all participants at the time of a cataract surgery.

In summary, from October 2016 to March 2018, 26 consecutive adults aged 60 years and over with a history of POAG and cataract were recruited from the Department of Ophthalmology, University Hospital of Angers, France, at the time of the cataract surgery. Parallel, 26 nonPOAG matched controls with cataract were recruited from the same Department at the time of the cataract surgery. Exclusion criteria for cases and controls were refusal to participate in research and native language other than French. Cases and controls participating in the study 4


Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


were included after having signed a non-objection form to the research. The study was conducted in accordance with the ethical standards set forth in the Helsinki Declaration (1983). The project was approved by the Medical Ethics Committee of the University Hospital of Angers, France (N° 2015-21).

The diagnosis of POAG was based on standard criteria, i.e. intraocular pressure > 21 mmHg at the time of diagnosis, glaucomatous optic nerve damage and excavation evidenced by ophthalmoscopy, and a thinner retinal nerve fiber layer (RNFL) measured by optical coherence tomography (OCT). All patients had open iridocorneal angles as determined by gonioscopic examination. Patients with secondary glaucoma, i.e. with signs of exfoliation or pigment dispersion following iritis, ocular trauma or intraocular surgery, angle-closure glaucoma, normal-tension glaucoma or isolated high intraocular pressure, were excluded. Standard automated perimetry (Humphrey Field Analyzer, Carl Zeiss, Dublin, CA, USA) with the 24-2 SITA Fast algorithm was performed on all POAG patients, and values of the visual field-mean defect (VF-MD) were used to grade the severity of POAG as “mild” with values lower than -6 dB, “moderate” with values between -6 dB and -12 dB, and “severe” with values higher than -12 dB, according to the perimetric Hoddap-Parrish-Anderson criteria. The reliability indices retained were false positive or false negative rates under 15%, and fixation losses under 20%. Other additional tests performed on patients with POAG included evaluation of the thickness of the RNFL using spectral domain optical coherence tomography (Cirrus OCT, Carl Zeiss Meditec, Dublin, CA, USA) and measurement of the central corneal thickness (CCT; Cirrus OCT, Carl Zeiss Meditec, Dublin, CA, USA). The best-corrected visual acuity was measured using the decimal Monoyer charts, with the results converted into logMAR units for statistical analysis. Intraocular pressures were measured using the

5



Page 6 of 30

Goldmann applanation tonometer. The current glaucoma treatment was documented for each patient.

The POAG group of patients (n = 26) was age-matched (to within 5 years) and sex-matched with a non-POAG control group (n = 26), all undergoing cataract surgery. The inclusion criteria of the control group were visual acuity ≥ 20/50 and the absence of any other associated ocular condition except cataract. The exclusion criteria were based on a family history of glaucoma, ocular hypertension or other intraocular pathologies, including retinal disorders. Aqueous humor samples were obtained from all patients at the beginning of the standardized phacoemulsification procedure of cataract surgery. The severity of the cataract and its visual impact was similar in both groups. All patients had fasted 12 hours prior to the procedure. Before surgery, all patients underwent an identical pre-operative protocol, including pupillary dilation with a dedicated topical insert 1 h before surgery consisting of tropicamide 0.28 mg and phenylphrine 5.4 mg (Mydriasert, Théa, Clermont-Ferrand, France). Standard disinfection with ophthalmic povidone iodine 5 % was carried out 2 minutes before beginning the procedure. Approximately 100-150 μL of aqueous humor were collected with a fine needle from each patient at the beginning of the surgical intervention. The aqueous humor samples were stored at -80°C until metabolomic analysis.

Metabolomic analysis Targeted quantitative metabolomic analysis was carried out with a Biocrates® Absolute IDQ p180 kit (Biocrates Life Sciences AG, Innsbruck, Austria) and a QTRAP 5500 mass spectrometer (SCIEX, Villebon-sur-Yvette, France), calibrated every three months, to quantify 188 metabolites, i.e. free carnitine and 39 acylcarnitines (C), 16 hexoses (H1), 21 amino acids, 21 biogenic amines, and 90 lipids. The lipid molecules quantified belong to four 6


Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


different classes: 14 lysophosphatidylcholines (lysoPC), 38 diacyl-phosphatidylcholines (PC aa), 38 acyl-alkyl-phosphatidylcholines (PC ae), and 15 sphingomyelins (SM). The full list of individual

metabolites

is

available

at

http://www.biocrates.com/products/research-

products/absoluteidq-p180-kit. High-performance liquid chromatography (HPLC) was used to separate individual members of the amino acid and biogenic amine classes before quantification with mass spectrometry. Metabolites from the other families were quantified after direct injection into the mass spectrometer by flow injection analysis coupled with tandem mass spectrometry (FIA-MS/MS).

All the reagents used in this analysis were of LC-MS grade and were purchased from VWR (Fontenay-sous-Bois, France) and Merck (Molsheim, France). Sample preparation and analyses were performed following the Kit User Manual. Briefly, each aqueous humor sample was thoroughly vortexed after thawing and centrifuged at 4°C for 5 minutes at 5000 g. Ten microliters of each sample were then added to the filter on the upper wells of the 96-well plate. Metabolites were extracted and derivatized for the quantitation of amino acids and biogenic amines. The extracts were finally diluted with MS running solvent before FIA and LC-MS/MS analysis. Three quality controls (QCs) composed of human plasma samples at three concentration levels: low (QC1), medium (QC2) and high (QC3), were used to evaluate the performance of the analytical assays. A seven-point serial dilution of the calibration compounds mixture was added to the plate of the kit to generate calibration curves for the quantification of amino acids and biogenic amines.

Statistical analyses Univariate analyses of clinical data were carried out using the bilateral Student’s t-test, with differences being considered significant at p≤ 0.05. 7



Page 8 of 30

Metabolites with more than 20% of concentration values below the lower limit of quantitation (LLOQ) or above the upper limit of quantitation (ULOQ) were excluded from the raw metabolomic data before statistical analyses.

Univariate analysis of the metabolomic data was performed using the non-parametric MannWhitney-Wilcoxon test (referred to below as Wilcoxon’s test) for quantitative variables. The Benjamini-Hochberg correction was used for comparing multiple metabolite concentrations to keep the false discovery rate below 5 %. Otherwise, the differences were considered significant at p≤ 0.05.

Unsupervised analysis was conducted using paired principal component analysis (pPCA) to detect similar groups of samples and outliers, i.e. samples displaying an atypical metabolite profile. Paired partial least squares – discriminant analysis (pPLS-DA) was performed on unit-variance to discriminate between POAG and control samples on the basis of their metabolic profiles. The data where divided into 18 training samples (~2/3) and 8 test samples (~ 1/3) from each group. The 1,562,275 models resulting from the allocation of 18 samples to the training set and 8 samples to the test set, out of a total of 26 samples, were tested. The predictive performance of the training models was evaluated using the area under the receiver operating characteristic (AUROC) and its associated p-value. This p-value measures the probability, for a given AUROC, of the corresponding model being the mean model, randomly predicting patient status, i.e. whether the patient belonged to the POAG group or the control group, with no additional information based on the metabolic profile. Since AUROCs and p-values cannot be considered to be normally distributed, median values instead of means were used to decide whether PLS models satisfactorily separate POAG from controls. Cut-off 8


Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


values of 0.8 and 0.05 were selected for the median AUROCs and p-values, respectively. The global performance of pPLS on our data was considered satisfactory only with the median AUROC value ≥ 0.8 and the median p-value ≤ 0.05. In this case, the selection of the variable was based on the variable importance for the projection (VIP) and the loading parameters. VIP values summarize the importance of each variable for the PLS-DA model, whereas the loading values are indicators of the relationship between the y vector containing the class information (i.e. whether POAG or control) and variables in the X matrix (i.e. metabolites). Variables with a VIP value greater than unity are important for group discrimination. Multivariate

analysis

was

performed

using

the

mixOmics

R

package

(http://www.Rproject.org).

RESULTS Clinical features of POAG patients and controls Comparison of demographic and environmental data as well as comorbid medical conditions between individuals with POAG (n = 26) and controls (n = 26) are presented in Table 1. The mean age of POAG patients did not differ significantly from that of controls, nor did the sex ratio. There was no difference regarding hypertension, hyperlipidemia, diabetes, thyroid disease, body mass index (BMI), kidney failure, or systemic medication. Ophthalmological examinations did not reveal any significant difference between the POAG group and controls regarding the intraocular pressure, which was improved by the glaucoma treatment for the POAG group (p = 0.13), and visual acuity (p = 0.71). Eighty-six percent of the POAG patients had a mild form of the disease, whereas 9% had a moderate form, and 5% a severe form of the disease.

9



Page 10 of 30

Table 1. Characteristics of the participants. Demographic data and comorbidity status, systemic medication, ophthalmological features and glaucoma medication of patients with POAG compared to controls. BMI: body mass index (weight/height2). IOP: intraocular pressure; CCT: central corneal thickness; RNFL: retinal nerve fiber layer; VF-MD: visual field mean defect. POAG Controls P(N = 26) (N = 26) values Demographic data and comorbidity Mean age (years) 74.69 74.92 0.87 Females (%) 46.15 46.15 1 2 Mean BMI (kg/m ) 27.11 27.23 0.93 Diabetes (%) 19.23 15.38 0.72 Hypertension (%) 57.70 61.54 0.78 Hyperlipidemia (%) 42.30 46.15 0.78 Thyroid disease (%) 15.38 15.38 1 Chronic kidney failure (%) 7.69 0.00 0.15 Systemic medications Antihypertensive drugs (%) 57.69 61.54 0.78 Lipid-lowering medications (%) 42.30 42.30 1 Antiplatelet therapy (%) 42.30 34.61 0.57 Oral diabetes medications (%) 15.38 11.53 0.69 Insulin (%) 0 3.84 0.32 Corticosteroids (%) 7.69 0 0.15 Thyroid hormone (%) 15.38 15.38 1 Estrogen (%) 0 0 1 Vitamin D (%) 7.69 7.69 1 Ophthalmological features and glaucoma medication Left eye (%) Right eye (%) Mean visual acuity (LogMar) Mean IOP (mmHg) Mean CCT (µm) Mean RNFL thickness (µm) Mean VF-MD (dB), Glaucoma severity (%) - Mild - Moderate - Severe Glaucoma medications (%) - Beta-blockers - Prostaglandin analogue - Alpha-2-agonists - Carbonic anhydrase inhibitor

1 50 50 0.06 15.88 558.22 69.25 -4.58

50 50 0.07 14.32 -

86.36 9.10 4.54

-

46.15 8461 11.53 26.92

-

1 1 0.71 0.13 10


Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Metabolomic analysis After validation of the experimental procedure based on quality control (QC) samples, 54 (28.7%) metabolites, accurately measured, were retained for statistical analysis: free carnitine, 3 acylcarnitines, the sum of hexoses composed at 90% from glucose, 20 amino acids, 16 biogenic amines and 13 lipids (these raw data are given in supplementary Table S1). Univariate analysis, after application of the Benjamini-Hochberg correction, showed that the concentration of 13 metabolites differed significantly between the POAG and the control groups, yielding a global false discovery rate of 0.045 (Figure 1). All these 13 metabolites as well as 9 other metabolites had median VIP values greater than 1 with multivariate analysis (Table 2).

11



Page 12 of 30

Table 2. Results of univariate and multivariate analyses for controls and POAG patients, ranked according to their VIP values. Median values together with the interquartile range (IQR) for each metabolite considered as important in multivariate and univariate analyses are indicated. The median value of the variable importance for the projection (VIP) and the adjusted P-value were used for ranking the importance of the metabolites. Acyl chain length and unsaturation degree in diacyl phosphatidylcholines (PC aa), lysophosphatidylcholine (Lyso PC a) and sphingomyeline (SM) are represented as X:Y where X represents the acyl chain length (for Lyso PC a and SM) or the sum of the two acyl-chain lengths (for PC aa) and Y represents the degree of unsaturation of the chains, respectively.

Metabolite

Median value [IQR] Controls POAG

VIP (median values)

Adjusted P-values

Creatinine

45.4 [12.3]

38.0 [13.8]

1.57

0.030

Propionylcarnitine (C3)

0.43 [0.10]

0.33 [0.10]

1.56

0.013

PC aa C34:1

0.20 [0.15]

0.10 [0.11]

1.55

0.028

Glutamine

677.0 [117.0]

580.0 [119.5]

1.54

0.029

Acetylcarnitine (C2)

3.74 [1.02]

3.05 [1.13]

1.52

0.029

Taurine

31.35 [11.75]

41.20 [19.30]

1.51

0.043

PC aa C36:2

0.080 [0.045]

0.045 [0.045]

1.38

0.03

PC aa C36:4

0.080 [0.045]

0.030 [0.028]

1.36

0.024

PC aa C38:4

0.051 [0.029]

0.037 [0.026]

1.33

0.025

SM C18:1

0.012 [0.009]

0.009 [0.005]

1.33

0.032

Carnitine (C0)

16.6 [6.0]

14.8 [5.2]

1.32

0.067

PC aa C32:1

0.026 [0.013]

0.016 [0.009]

1.30

0.039

PC aa C34:2

0.110 [0.074]

0.056 [0.055]

1.30

0.031

Trans-4-OH-Proline

3.95 [3.25]

3.46 [1.71]

1.27

0.031

Isoleucine

85.8 [22.4]

74.6 [24.0]

1.22

0.063

PC aa C30:2

0.004 [0.002]

0.002 [0.003]

1.17

0.091

Alanine

285.0 [69.0]

251.0 [79.8]

1.13

0.114

Acetylornithine

0.35 [0.56]

0.16 [0.33]

1.12

0.111

Butyrylcarnitine (C4)

0.18 [0.06]

0.15 [0.04]

1.11

0.087

Lyso PC a C28:1

0.023 [0.009]

0.018 [0.008]

1.10

0.129 12


Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Leucine

108.0 [28.6]

96.9 [23.3]

1.08

0.095

Glycine

26.9 [17.8]

21.95 [7.55]

1.07

0.147

Figure 1. Results of univariate analysis. After application of the Benjamini-Hochberg correction, univariate analysis revealed significant differences of metabolite concentration in the aqueous humor of POAG patients compared to controls. The paired PCA scatter plot of metabolomic data showed no spontaneous grouping according to the glaucomatous condition, nor any strong outliers in the first and second principal plans (PC1 vs. PC2) (Figure 2).

13



Page 14 of 30

Figure 2. Paired PCA scatter plot. The scatter plot constructed from the matrix of metabolites for the 26 POAG patients (red circles) and their paired controls (blue circles) showed no clear grouping nor any outliers (according to Hotelling's T2 range). PC1, PC2: principal components 1 and 2. Supervised, paired PLS-DA showed a good predictive performance with median AUROC values of 0.89 for the test sets with p-values of 0.0087 (Figure S1).

Using median coordinates for the training sets, the scatter plot of the two latent variables of the pPLS-DA models shows good discrimination between POAG patients and controls (Figure 3).

Figure 3. Paired PLS-DA scatter plot. The paired PLS-DA scatter plot shows a clear distinction between glaucomatous patients (red dots) and their paired controls (blue dots). Projecting samples on both axes shows that LV1 carried out most of the between-group discrimination with only two paired samples not been well predicted (red/blue dots in the upper-left/lower-right quadrants). LV1, and LV2: Latent variables 1 and 2, respectively.

Median VIP and loading estimations, or the “volcano plot”, for all these models are shown in Figure 4. Compared to controls, POAG samples had two metabolites with VIP > 1, taurine and spermine, at lower concentrations and 21 metabolites at higher concentrations: 14


Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


creatinine,

carnitine,

three

acylcarnitines

(acetylcarnitine,

propionylcarnitine,

butyrylcarnitine), seven amino acids (glutamine, glycine, alanine, leucine, isoleucine, acetylornithine and trans-4-OH-Proline), seven phosphatidylcholines, one lysophosphatidylcholine and one sphingomyelin.

Figure 4. Volcano plot (Loading versus VIP). The volcano plot was constructed using median loading and VIP values from 1,562,275 possible models. Only the most important metabolites, i.e. those with VIP ≥ 1, have been labelled. Negative loadings indicate increased metabolite concentrations in the aqueous humor from POAG patients compared to controls. The metabolomic signature of glaucoma in the aqueous humor is associated with the decreased level of taurine and the increased levels of three biogenic amines: creatinine, trans4-hydroxyproline (t4-OH-Pro) and acetyl-ornithine (Ac-Orn); five amino acids: glutamine (Gln), glycine (Gly), alanine (Ala), leucine (Leu) and isoleucine (Ile); free carnitine and three acylcarnitines: acetylcarnitine (C2), propionylcarnitine (C3), butyrylcarnitine (C4); one lysophosphatidylcholine: lyso PC a 28:1; seven unsaturated diacylphosphatidylcholines: PC aa 30:2, 32:1, 34:1, 34:2, 36:2, 36:4 and 38:2 and one sphingomyelin: SM 18:1. Levels of spermine were also decreased in the aqueous humor of POAG patients, but its VIP value is just short of 1. Biogenic amines, amino acids, carnitine and acylcarnitines, lysophosphatidylcholines, phosphatidylcholines and sphingomyelins are represented in pink, light green, brown, blue, orange and yellow bubbles, respectively. Red-rimmed bubbles indicate metabolites that differ significantly between POAG patients and controls in univariate analysis (Wilcoxon’s test) after the Benjamini-Hochberg correction (see Table 2).

15



Page 16 of 30

Thus, among the 188 polar metabolites and lipids analyzed by the Biocrates® Absolute IDQ p180 kit, 54 were accurately measured according to the quality controls supplied with the kit. Multivariate analysis showed that 23 of the metabolites could discriminate between POAG patients and controls. Thirteen of these 23 metabolites were also found discriminant using univariate analysis after correction for the false discovery rate. Compared to controls, the metabolomic profile of the aqueous humor of POAG patients includes higher concentrations of nine phospholipids, seven amino acids, three short-chain acylcarnitines, free carnitine and creatinine, together with lower concentrations of taurine and spermine. Some of these features have already been described in other human and animal studies, attesting to the reliability of our biochemical and statistical pipelines, whereas the other findings are new. Indeed, increased concentrations of trans-4-OH-Proline, alanine, glutamine, creatinine and fatty acids were also found in the aqueous humor of a rat glaucoma model with increased intraocular pressure generated by the injection of sodium hyaluronate.8

Metabolic signature interpretation A higher concentration of hydroxyproline in the aqueous humor has already been reported in patients with POAG14 and with the pseudoexfoliation syndrome, the leading cause of secondary glaucoma.15 Since hydroxyproline is produced only by the hydrolysis of collagen, the former observation suggests that the increased collagen turnover in tissues drained by the aqueous humor may contribute to the pathogenesis of glaucoma. Indeed, higher concentrations of hydroxyproline and abnormal collagen content and composition have been reported in the lamina cribosa from necropsy cases of glaucoma.16 However, the trabecular meshwork was not involved in this process, since no difference of collagen and hydroxyproline content was found in this tissue between glaucoma patients and controls.17

16


Page 17 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The decreased concentration of spermine has never been reported in the aqueous humor of POAG patients. Polyamines, such as putrescine, spermine and spermidine, which are ubiquitous poly-cations supplied by diet and endogenous biosynthesis, bind to nucleic acids in order to stabilize them. Involved in the processes of gene expression, cell growth and proliferation, these molecules possess anti-apoptotic and neuroprotective properties. Interestingly, we previously found that the concentration of spermine was also reduced in the plasma of patients with POAG and in patients with the exfoliation syndrome.18,19 The decrease of this polyamine in the aqueous humor may play an important role in the pathogenesis of POAG, since the spermine-related spermidine has been reported to protect the optic nerve in a mouse model of normal-tension glaucoma.20 Since the concentration of polyamines declines with age in many species, we hypothesized that the age-related decrease of spermine and spermidine may contribute to the vulnerability of retinal ganglion cells, as already observed in the plasma from glaucoma patients.18

Taurine, a sulfonic amino acid in which the carboxylic group is replaced by a sulfonic acid group, is provided mainly through the diet and to a lesser extent by endogenous biosynthesis. It is one of the most abundant amino acids in the retina, cornea and lens, as well as in the vitreous and aqueous humors.21 Taurine has an important cytoprotective function against oxidative stress, inflammation and osmoprotection, and its concentration decreases with age, in particular in the cornea and lens.22 In addition, taurine ensures retinal ganglion cell survival in a rat model of glaucoma23 and in albino rats, taurine depletion was shown to increase retinal ganglion cell loss.24 Provided by drinking water, taurine protects the optic nerve in the DBA/2J mouse model of glaucoma with increased intraocular pressure.21 Surprisingly, in a canine model of glaucoma,25 the plasma level of taurine was found to be higher in affected dogs than in controls but this was attributed to a compensatory mechanism counteracting 17



Page 18 of 30

oxidative stress. Thus, in the line with animal data, the significant reduction of taurine concentration, reported here for the first time in the aqueous humor of POAG patients, might jeopardize retinal ganglion cell survival. Furthermore, since taurine controls the osmoregulation of the corneal epithelium thereby preventing hypertonicity, the drastic reduction of this amino acid may play a central role in POAG etiology.

Higher amino acid levels than in controls have been reported in the aqueous humor of POAG patients.26 Similarly, seven amino acids (glutamate, glutamine, glycine, alanine, serine and arginine) had higher than normal concentrations in the retina of 11-month-old DBA/2J model mice with glaucoma.27 In line with these findings, our study has also found increased concentrations of seven amino acids (glutamine, glycine, alanine, leucine, isoleucine, hydroxyl-proline and acetyl-ornithine) in the aqueous humor of patients with POAG.

The increased concentration of free carnitine and three acylcarnitines with short-chain fatty acids (C2, C3, C4) has never so far been reported in the aqueous humor of POAG patients. Importantly, carnitine has been shown to have neuroprotective, anti-apoptotic and antioxidant properties in the retinal cells of an experimental mouse model of high intraocular pressure.28 Thus, increasing the concentration of carnitine may protect retinal cells against disease-related stress such as that due to the production of reactive oxygen species. In addition, the absence of long-chain acylcarnitines in our signature does not argue in favor of a defect of mitochondrial oxidation in the cells drained by the aqueous humor. Conversely, the increased concentration of six amino acids together with that of short-chain acylcarnitines, which are associated with the degradation of amino acids, is compatible with a deficient amino acid metabolism in cells drained by the aqueous humor. Interestingly, we also previously found that the plasma concentration of C3 and C4 acylcarnitines was higher in POAG patients than 18


Page 19 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


in controls,18 thus reinforcing the systemic role of these metabolites in the pathogenesis of glaucoma.

Increased concentrations of creatinine have been found in the aqueous humor of a rat model of glaucoma8 but have never been reported in POAG patients. The symmetrical positions of taurine and creatinine in the volcano plot, reflecting their opposite variation of concentration, suggests that creatinine plays an osmolytic role compensating the decreased level of taurine. Creatinine is a catabolic product of creatine, which was not measured in our study. However, we may note that, in muscle cells, creatine acts similarly to well-established osmolytes, such as betaine, taurine and inositol, protecting these cells against hypertonic stress,29 and creatine is also believed to act as an osmolyte in the human brain.30 As for taurine, it is possible that the treatment against intraocular pressure may be responsible for the deregulation of this osmolyte, although the spontaneously increased concentration of creatinine found in a glaucoma rat model8 does not argue in favor of this hypothesis.

A study comparing the hypertensive to the hypotensive state of the DBA/2J mouse model showed that phospholipids were globally decreased, without any discernable differences in the concentration of phosphatidylcholines.10 Phospholipid profiles of the aqueous humor were also significantly different between POAG patients and controls,12 but in contrast to the mouse model, the concentration of most of the diacylglycerophosphocholines and 14 sphingomyelins were increased. Our study also revealed the increased concentration of 7 phosphatidylcholines, which were all diacyl forms of the unsaturated phosphatidylcholines, as well as one sphingomyelin and one lysophosphatidylcholine. This phospholipid remodeling reflects modifications of the membranes from the cells drained by the aqueous humor. Interestingly, we previously reported the significantly increased concentration of four of the 19



Page 20 of 30

76 phosphatidylcholines in the plasma of POAG patients compared to controls, with two compounds shared by the plasma and the aqueous humor (PC aa 34:2 and PC aa 36:4), thus highlighting these metabolites as biomarkers of glaucoma.18

This alteration of the concentration of some metabolites in the aqueous humor of patients affected by glaucoma, despite the therapeutic normalization of their intraocular pressure, suggests that the disease affects the permeability of the blood-aqueous barrier. Potentially, there are several pathophysiological causes for such an alteration of the blood-aqueous barrier in glaucoma, e.g. vascular injury, inflammation, and metabolic remodeling. As has been shown for proteins, this barrier is highly selective, whether in its physiological state or when it is altered.31 Thus, this selectivity of the blood-aqueous barrier could also contribute to the content of the metabolic signature.

Limitations The main limitations of this study are due to the small number of participants, all from a single center, and the relatively few metabolites investigated. A pre-concentration of the aqueous humor before quantification would have been useful to augment the number of metabolites accurately measured. However, the quantitative nature of our targeted metabolomic approach and the diversity of the metabolic pathways explored revealed significant metabolic changes already identified in other studies in humans or animals, thus confirming the relevance of our work. Another limitation of the study arises from the difficulty of distinguishing the metabolic changes due to the glaucoma itself from those due to topical glaucoma treatments. However, the metabolic features reported here are consistent with those found in animal models, thus suggesting a link to glaucoma rather than to the treatment of the disease. Lastly, the lipid measurements with the AbsoluteIDQ® p180 Kit 20


Page 21 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


were carried out employing flow injection analysis tandem mass spectrometry (FIA-MS/MS), quantified by an internal standard. The FIA-MS/MS measurements cannot differentiate between the fatty acids linked to the glycerol backbone or define their bond type. The use of Ion mobility mass spectrometry, or a separation technique such as the HPLC retention time combined with the m/z ratios, would have provided more structural information on the lipids detected in the current study.

CONCLUSIONS The metabolomic signature of aqueous humor from patients with POAG reveals four main features: 1) the modified concentration of two metabolites involved in osmoprotection, i.e. taurine and creatinine; 2) the reduced concentration of two strong neuroprotectors, i.e. spermine and taurine, together with the increased concentration of carnitine; 3) the remodeled amino acid metabolism involving seven amino acids and three acylcarnitines; and 4) the structurally remodeled membranes of cells drained by the aqueous humor (hydroxyproline and phospholipids). In addition, some metabolic modifications of spermine, C3 and C4 acylcarnitines, PC aa 34:2 and PC aa 36:4, reproduce metabolomic features already found in the plasma from POAG patients, reinforcing their importance in the pathogenesis of glaucoma. Finally, our study suggests that the supplementation of some of these metabolites, notably taurine, spermine and carnitine, may lead to protective outcomes in glaucoma. ASSOCIATED CONTENT A table showing the raw data of the 54 metabolites accurately measured in the aqueous humor of POAG patients and controls (Supplementary Table S1). Figure S1. Predictive performance of the pPLS-DA models on the test sets. AUTHOR INFORMATION Corresponding author *Email: [email protected] Pascal Reynier: Orcid number: 0000-0003-0802-4608 21



Page 22 of 30

Author Contributions CA, DM, PR and JMCB conceived of the experimental plan, AB, PG, SL, JM and PAB performed experiments, GL, DB, GS, VP, CA, DM, PR and JMCB analyzed data and wrote the manuscript. All authors have given approval to the final version of the manuscript. Funding This work was supported by grants from the University Hospital of Angers and from the UNADEV (Union Nationale des Aveugles et Déficients Visuels), in partnership with the ITMO NNP/AVIESAN (Alliance Nationale pour les Sciences de la Vie et de la Santé). Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS We are grateful to the individuals participating in this study, to Kanaya Malkani for critical reading and comments on the manuscript. We acknowledge support from the Institut National de la Santé et de la Recherche Médicale (INSERM), the Centre National de la Recherche Scientifique (CNRS), the University of Angers, and the University Hospital of Angers. We also wish to thank the following patients’ associations for their support: “Fondation VISIO”, “Ouvrir les Yeux”, “Union Nationale des Aveugles et Déficients Visuels”, “Retina France”, and “Kjer France”. REFERENCES 1) Pietrowska, K.; Dmuchowska, D. A.; Samczuk, P.; Kowalczyk, T.; Krasnicki, P.; Wojnar, M.; Skowronska, A.; Mariak, Z.; Kretowski, A.; Ciborowski, M. LC-MS-Based Metabolic Fingerprinting of Aqueous Humor. J. Anal. Methods Chem. 2017, 2017, 6745932. 2) Pietrowska, K.; Dmuchowska, D. A.; Krasnicki, P.; Bujalska, A.; Samczuk, P.; Parfieniuk, E.; Kowalczyk, T.; Wojnar, M.; Mariak, Z.; Kretowski, A.; Ciborowski, M. An exploratory LC-MS-based metabolomics study reveals differences in aqueous humor composition between diabetic and non-diabetic patients with cataract. Electrophoresis 2018, 39, 12331240. 3) Tessem, M. B.; Bathen, T. F., Cejková, J.; Midelfart, A. Effect of UV-A and UV-B irradiation on the metabolic profile of aqueous humor in rabbits analyzed by 1H NMR spectroscopy. Invest. Ophthalmol. Vis. Sci. 2005, 46, 776-81. 4) Wang, HY.; Wang, Y.; Zhang, Y.; Wang, J.; Xiong, S. Y.; Sun, Q. Crosslink between lipids and acute uveitis: a lipidomic analysis. Int. J. Ophthalmol. 2018, 11, 736746. 5) Ji, Y.; Rong, X.; Lu, Y. Metabolic characterization of human aqueous humor in the cataract progression after pars plana vitrectomy. BMC Ophthalmol. 2018, 18, 63. 6) Barbas-Bernardos, C.; Armitage, E. G.; García, A.; Mérida, S.; Navea, A.; Bosch-Morell, F.; Barbas, C. Looking into aqueous humor through metabolomics spectacles - exploring 22


Page 23 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


its metabolic characteristics in relation to myopia. J. Pharm. Biomed. Anal. 2016, 127, 1825. 7) Snytnilova, O. A.; Yanshole, L.; Iskakov, I. A.; Yanshole, V. V.; Chernykh, V. V.; Stepakov, D. A.; Novoselov, V. P.; Tsentalovich, Y. P. Quantitative metabolomic analysis of the human cornea and aqueous humor. Metabolomics 2017, 13, 152. 8) Mayordomo-Febrer, A; López-Murcia, M; Morales-Tatay, J. M; Monleón-Salvado, D.; Pinazo-Durán, M. D. Metabolomics of the aqueous humor in the rat glaucoma model induced by a series of intracamerular sodium hyaluronate injection. Exp. Eye Res. 2015, 131, 84-92. 9) Williamson, B. K.; Hawkey, N. M.; Blake, D. A.; Frenkel, J. W.; McDaniel, K. P.; Davis, J. K.; Satija, C.; Beazer, A.; Dhungana, S.; Carlson, J.; McRitchie, S.; Ayyala, R. S. The Effects of Glaucoma Drainage Devices on Oxygen Tension, Glycolytic Metabolites, and Metabolomics Profile of Aqueous Humor in the Rabbit. Transl. Vis. Sci. Technol. 2018, 7, 14. 10) Wang, H.; Edwards, G.; Garzon, C.; Piqueras, C.; Bhattacharya, S. K. Aqueous humor phospholipids of DBA/2J and DBA/2J-Gpnmb(+)/SjJ mice. Biochimie 2015, 113, 59-68. 11) Edwards, G.; Aribindi, K.; Guerra, Y.; Lee, R. K.; Bhattacharya, S. K. Phospholipid profiles of control and glaucomatous human aqueous humor. Biochimie 2014, 101, 232-47. 12) Cabrerizo, J.; Urcola, J. A.; Vecino, E. Changes in the Lipidomic Profile of Aqueous Humor in Open-Angle Glaucoma. J. Glaucoma 2017, 26, 349-355. 13) Aljohani, A. J.; Munguba, G. C.; Guerra, Y.; Lee, R. K.; Bhattacharya, S. K. Sphingolipids and ceramides in human aqueous humor. Mol. Vis. 2013, 19, 1966-84. 14) Ghanem, A. A.; Mady, S. M.; El awady, H. E.; Arafa, L. F. Homocysteine and hydroxyproline levels in patients with primary open-angle glaucoma. Curr. Eye Res. 2012, 37, 712-8. 15) Yagci, R.; Ersöz, I.; Aydin, B.; Beyaz, E.; Gürel, A.; Durmuş, M.; Duman, S. Aqueous humor and serum concentration of hydroxyproline in pseudoexfoliation syndrome. J. Glaucoma 2007, 16, 225-9. 16) Tengroth, B.; Ammitzbøll, T. Changes in the content and composition of collagen in the glaucomatous eye--basis for a new hypothesis for the genesis of chronic open angle glaucoma--a preliminary report. Acta Ophthalmol. (Copenh.) 1984, 62, 999-1008. 17) Finkelstein, I.; Trope, G. E.; Basu, P. K.; Hasany, S. M.; Hunter, W. S. Quantitative analysis of collagen content and amino acids in trabecular meshwork. Br. J. Ophthalmol. 1990, 74, 280-2. 18) Leruez, S.; Bresson, T.; Chao de la Barca, J. M.; Marill, A.; de Saint Martin, G, Buisset.; A, Muller, J.; Tessier, L.; Gadras, C.; Verny, C.; Amati-Bonneau, P.; Lenaers, G.; 23



Page 24 of 30

Gohier, P.; Bonneau, D.; Simard, G.; Milea, D.; Procaccio, V.; Reynier, P. The metabolomic signature of glaucoma points to mitochondrial dysfunction, senescence and polyamines deficiency. Invest. Ophthalmol. Vis. Sci. 2018, 59, 4355-4361. 19) Leruez, S.; Marill, A.; Bresson, T.; de Saint Martin, G.; Buisset, A.; Muller, J.; Tessier, L.; Gadras, C.; Verny, C.; Gohier, P.; Amati-Bonneau, P.; Lenaers, G.; Bonneau, D.; Simard, G.; Milea, D.; Procaccio, V.; Reynier, P.; Chao de la Barca, J. M. A plasma metabolomic signature of the exfoliation syndrome involves amino acids, acyl-carnitines and polyamines. Invest. Ophthalmol. Vis. Sci. 2018, 59, 1025-1032. 20) Noro, T.; Namekata, K.; Azuchi, Y.; Kimura, A.; Guo, X.; Harada, C.; Nakano, T.; Tsuneoka, H.; Harada, T. Spermidine Ameliorates Neurodegeneration in a Mouse Model of Normal Tension Glaucoma. Invest. Ophthalmol. Vis. Sci. 2015, 56, 5012-9. 21) Froger, N.; Jammoul, F.; Gaucher, D.; Cadetti, L.; Lorach, H.; Degardin, J.; Pain, D.; Dubus, E.; Forster, V.; Ivkovic, I.; Simonutti, M.; Sahel, J. A.; Picaud, S. Taurine provides neuroprotection against retinal ganglion cell degeneration. PLoS One 2012, 7, e42017. 22) Wang, Y.; Grenell, A.; Zhong, F.; Yam, M.; Hauer, A.; Gregor, E.; Zhu, S.; Lohner, D.; Zhu, J.; Du, J. Metabolic signature of the aging eye in mice. Neurobiol. Aging 2018, 71, 223-233. 23) Froger, N.; Jammoul, F.; Gaucher, D.; Cadetti, L.; Lorach, H.; Degardin, J.; Pain, D.; Dubus, E.; Forster, V.; Ivkovic, I.; Simonutti, M.; Sahel, J. A.; Picaud S. Taurine is a crucial factor to preserve retinal ganglion cell survival. Adv. Exp. Med. Biol. 2013, 775, 6983. 24) García-Ayuso, D.; Di Pierdomenico, J.; Hadj-Said, W.; Marie, M.; Agudo-Barriuso, M.; Vidal-Sanz, M.; Picaud, S.; Villegas-Pérez, M. P. Taurine Depletion Causes ipRGC Loss and Increases Light-Induced Photoreceptor Degeneration. Invest. Ophthalmol. Vis. Sci. 2018, 59, 1396-1409. 25) Boillot, T.; Rosolen, S. G.; Dulaurent, T.; Goulle, F.; Thomas, P.; Isard, P. F.; Azoulay, T.; Lafarge-Beurlet, S.; Woods, M.; Lavillegrand, S.; Ivkovic, I.; Neveux, N.; Sahel, J. A.; Picaud, S.; Froger, N. Determination of morphological, biometric and biochemical susceptibilities in healthy Eurasier dogs with suspected inherited glaucoma. PLoS One 2014, 9, e111873. 26) Hannappel, E.; Pankow, G.; Grassl, F.; Brand, K.; Naumann, G. O. Amino acid pattern in human aqueous humor of patients with senile cataract and primary openangle glaucoma. Ophthalmic Res. 1985, 17, 341-3. 27) Schuettauf, F.; Thaler, S.; Bolz, S.; Fries, J.; Kalbacher, H.; Mankowska, A.; Zurakowski, D.; Zrenner, E.; Rejdak, R. Alterations of amino acids and glutamate transport in the DBA/2J mouse retina; possible clues to degeneration. Graefes Arch. Clin. Exp. Ophthalmol. 2007, 245, 1157-68.

24


Page 25 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


28) Calandrella, N.; De Seta, C.; Scarsella, G.; Risuleo, G. Carnitine reduces the lipoperoxidative damage of the membrane and apoptosis after induction of cell stress in experimental glaucoma. Cell Death Dis. 2010, 1, e62. 29) Alfieri, R. R.; Bonelli, M. A.; Cavazzoni, A.; Brigotti, M.; Fumarola, C.; Sestili, P.; Mozzoni, P.; De Palma, G.; Mutti, A.; Carnicelli, D.; Vacondio, F.; Silva, C.; Borghetti, AF.; Wheeler, K. P.; Petronini, P. G. Creatine as a compatible osmolyte in muscle cells exposed to hypertonic stress. J. Physiol. 2006, 576, 391-401. Videen, J. S.; Michaelis, T.; Pinto, P.; Ross, B. D. Human cerebral osmolytes during chronic hyponatremia. A proton magnetic resonance spectroscopy study. J. Clin. Invest. 1995, 95, 788-93.

30)

Liu, P., Thomson, B.R., Khalatyan, N., Feng, L., Liu, X., Savas, J.N., Quaggin, S.E., Jin, J. Selective permeability of mouse blood-aqueous barrier as determined by 15Nheavy isotope tracing and mass spectrometry. Proc. Natl. Acad. Sci. U S A. 2018, 115, 9032-9037.

31)

25



Page 26 of 30

For TOC ONLY:

26


Page 27 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1 338x190mm (96 x 96 DPI)





Page 28 of 30

Page 29 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60







Page 30 of 30

Metabolomic profiling of aqueous humor in glaucoma points to taurine

Recommend Documents