Next Article in Journal
The Effects of Cannabidiol on Aqueous Humor Outflow and Trabecular Meshwork Cell Signaling
Next Article in Special Issue
Systematic Interrogation of the Temperature Perturbation in the Insulin Signaling Pathway for Optogenetic Stimulation
Previous Article in Journal
The Wheat Head Blight Pathogen Fusarium graminearum Can Recruit Collaborating Bacteria from Soil
Previous Article in Special Issue
Mechanical Stretch Induced Osteogenesis on Human Annulus Fibrosus Cells through Upregulation of BMP-2/6 Heterodimer and Activation of P38 and SMAD1/5/8 Signaling Pathways
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Cells
Volume 11
Issue 19
10.3390/cells11193005
cells-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Metabolomics in Diabetic Retinopathy: From Potential Biomarkers to Molecular Basis of Oxidative Stress

by
Qizhi Jian
1,2,3,4,
Yingjie Wu
5,6,7,* and
Fang Zhang
1,2,3,4,*
1
National Clinical Research Center for Eye Diseases, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200080, China
2
Shanghai Key Laboratory of Ocular Fundus Diseases, Shanghai 200080, China
3
Shanghai Engineering Center for Visual Science and Photomedicine, Shanghai 200080, China
4
Shanghai Engineering Center for Precise Diagnosis and Treatment of Eye Diseases, Shanghai 200080, China
5
Institute for Genome Engineered Animal Models of Human Diseases, National Center of Genetically Engineered Animal Models for International Research, Liaoning Provence Key Laboratory of Genome Engineered Animal Models, Dalian Medical University, Dalian 116000, China
6
Shandong Provincial Hospital, School of Laboratory Animal & Shandong Laboratory Animal Center, Science and Technology Innovation Center, Shandong First Medical University & Shandong Academy of Medical Sciences, Jinan 250021, China
7
Department of Molecular Pathobiology, New York University College of Dentistry, New York, NY 10010, USA
*
Authors to whom correspondence should be addressed.
Cells 2022, 11(19), 3005; https://doi.org/10.3390/cells11193005
Submission received: 31 July 2022 / Accepted: 22 September 2022 / Published: 26 September 2022
(This article belongs to the Special Issue Molecular Mechanism of Stress, Stress Response, and Adaptation)
Browse Figures
Versions Notes

Abstract

:
Diabetic retinopathy (DR), the leading cause of blindness in working-age adults, is one of the most common complications of diabetes mellitus (DM) featured by metabolic disorders. With the global prevalence of diabetes, the incidence of DR is expected to increase. Prompt detection and the targeting of anti-oxidative stress intervention could effectively reduce visual impairment caused by DR. However, the diagnosis and treatment of DR is often delayed due to the absence of obvious signs of retina imaging. Research progress supports that metabolomics is a powerful tool to discover potential diagnostic biomarkers and therapeutic targets for the causes of oxidative stress through profiling metabolites in diseases, which provides great opportunities for DR with metabolic heterogeneity. Thus, this review summarizes the latest advances in metabolomics in DR, as well as potential diagnostic biomarkers, and predicts molecular targets through the integration of genome-wide association studies (GWAS) with metabolomics. Metabolomics provides potential biomarkers, molecular targets and therapeutic strategies for controlling the progress of DR, especially the interventions at early stages and precise treatments based on individual patient variations.

1. Introduction

Diabetic retinopathy (DR) is a major complication of diabetes mellitus (DM), and one of the leading causes of vision impairment and blindness in working-age adults globally [1,2,3,4]. In 2030, the number of adults worldwide with DR is estimated to be 129.84 million, and the number is projected to increase to 160.50 million in 2045 [5]. The economic burden increased accordingly.
Although important advances have been made in the diagnosis and treatment of DR in the past few decades, more effective diagnostic markers and therapeutic strategies are still lacking. Hemoglobin A1c (HbA1c) for monitoring the levels of glucose is the validated systemic biomarker of DR [6], while the differences in the level of HbA1c explained only 6.6% of the alteration in the risk of DR for the entire study cohort in a diabetes control and complications trial [7,8]. The worsening of DR, including the early-phase non-proliferative diabetic retinopathy (NPDR) and advanced-phase proliferative diabetic retinopathy (PDR), is associated with the initiation of effective treatment of glycaemia in patients with diabetes. Moreover, there are few measures available for early intervention in DR beyond regulating hyperglycemia and hypertension, preventing dyslipidemia, and cessation of tobacco smoking [9]. All the treatments including anti-vascular endothelial growth factor (anti-VEGF) therapy and laser photocoagulation are focused on the PDR and carry sight-threatening effects [9,10,11]. Thus, there is still an urgent need to identify novel biomarkers and effective therapeutic strategies to screen and treat the progression of DR.
Research progress supports that oxidative stress, caused by the disruption of redox balance, is closely related to metabolic dysregulation in the retina and is a key contributor to the pathogenesis of DR [12,13]. Metabolism-based regulation of oxidative stress would be a potential target for developing promising treatments for DR.
Metabolomics enables the detailed characterization of metabolic phenotypes and metabolic derangements that underlie diseases. It could afford the discovery of biomarkers and new therapeutic targets that may be used to either diagnose disease or monitor the activity of therapeutics [14]. The application of metabolomics in DR has systematically reflected abnormal metabolic changes by comparing the small molecule composition of various ocular and blood samples from DR patients, mammals, tissues, and cells [15]. In this manuscript, we aim to review the latest advances in metabolomics of DR and summarize potential biomarkers and molecular targets for DR by integrating metabolomics with genome-wide association studies (GWAS).

2. Overview of Metabolomics of DR

In recent years, an increasing number of studies have applied metabolomics in DR. We retrieved relevant articles by searching PubMed before 28 February 2022, with the following search method: (“metabolomics” or “lipidomics” or “metabonomics” or “metabolome” or “metabolic profiling”) AND “diabetic retinopathy”. Thirty-nine original articles were finally determined and the details of these articles including species, samples, platforms, and differential metabolites are summarized in Figure 1 and Table 1.
According to the search results, metabolomics studies of DR using various biological samples have become widely available since 2009 (Figure 1). In 2009, Abhary et al. performed metabolomics profiling of serum from patients with DR and found that the serum levels of L-arginine, asymmetric dimethylarginine (ADMA), and symmetric dimethylarginine (SDMA) were increased in DR patients compared to diabetic patients without DR [26]. In the same year, Young et al. used metabolomic analysis of human vitreous humor to differentiate ocular inflammatory diseases including proliferative vitreoretinopathy (PVR) and proliferative diabetic retinopathy, and showed that PVR and PDR could be separated by the metabolomic analysis of vitreous humor [39]. In 2011, Li et al. analyzed the metabolome of plasma from patients with DR and identified pyruvic acids, l-aspartic acid, β-hydroxybutyric acid, methylmaleic acid, citric acid, glucose, stearic acid, trans-oleic acid, linoleic acid, and arachidonic acid as differential metabolites [16]. Koehrer et al. identified the metabolic profiles of erythrocytes in DR patients and found that the levels of docosahexaenoic acid, arachidonic acid, and PUFAs in red blood cells were decreased in 2014 [32]. The metabolic profile of aqueous humor from patients with DR was measured by Kunikata et al. in 2017 [37]. They identified cysteine persulfides, oxidized glutathione trisulfide (GSSSG) and cystine were changed in aqueous humor. The next year, Sas et al. examined the lipidome in plasma and retinal tissues using a mouse model of type 2 diabetes with microvascular complications and found alterations of 15 lipids in both plasma and retina tissues [52]. Lin et al. investigated the metabolomic profile of type 2 diabetes in cerebrospinal fluid and identified that alanine, histidine, leucine, pyruvate, tyrosine, and valine showed the potential of biomarkers for DR in 2019 [47]. In 2021, Ye and Zhou respectively collected stool samples from DR patients to analyze the alterations of gut metabolomes linking DR to the gut metabolome—gut-retina-axis [33,34]. Meanwhile, Quek et al. analyzed the urinary metabolic profiles of DR patients and found the alterations of citrate, ethanolamine, formate, and hypoxanthine in urine [50].
From the perspective of species, 35 studies recruited patients with DR for metabolomic analysis, and the other four studies selected diabetic animal models including mice, rats and zebrafish. There is a wide selection of biofluids for metabolomic analysis in human studies, including circulating blood (plasma and serum), eye fluids (vitreous humor and aqueous humor), and other samples (retinal tissues, stools, urines, red blood cells, and cerebrospinal fluids). Different types of samples have their own characteristics and advantages. Circulating blood, due to its easier availability and lower invasiveness, is the most commonly used sample and can provide a global metabolomic picture [55]. Both serum and plasma can be obtained from blood, and the main difference between them is the presence or absence of clotting factors [56]. In terms of impact on metabolite detection, plasma appears to have better reproducibility, with serum having higher concentrations [57]. Eye fluids including vitreous humor and aqueous humor can directly reflect intraocular metabolic variations. However, the vitreous humor, a highly aqueous eye fluid interfacing with the retina, can only be obtained from subjects with PDR during surgery such as a vitrectomy, which results in the absence of vitreous samples of NPDR. Tears can be obtained non-invasively and can also reflect the conditions of the oculi posterior segment [6]. However, to our knowledge, tear metabolomics have not been applied to the study of DR. In addition, stool samples can reflect alterations of fecal metabolome and gut microbiota composition, linking DR to the gut metabolome and microbiota—gut-retina-axis [58].
From the perspective of the metabolomics analysis platform, there are two main tools: nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS). Thirty-three studies used MS for metabolite analysis and five studies used NMR. NMR spectroscopy can be applied to biological samples in various states including liquid, solid, and gaseous samples [59]. The proton NMR method is the most widely applied NMR technique [60]. A significant advantage of NMR is the small number of samples required [15]. MS is often used in tandem with liquid chromatography (LC) or gas chromatography (GC), which are techniques applied to separate metabolites. In particular, LC-MS has been widely used in recent years. MS has far better sensitivity than NMR, allowing it to measure a wider spectrum of metabolites [15]. Overall, the use of NMR and MS has greatly facilitated the development of metabolomics.

3. Potential Metabolomics Biomarkers of DR

Metabolomics has been utilized extensively for the identification of single metabolites and their use as biomarkers [61]. In DR research, 14 studies applied explicit statistical methods to identify new metabolomics biomarkers and evaluate the performance for disease diagnosis of biomarker models as listed in Table 2. These studies primarily covered the human serum, plasma, vitreous humor, aqueous humor, cerebrospinal fluid, and stool. Human plasma is the most widely used sample for identifying biomarkers of DR, and various metabolites in human plasma have been reported to have biomarker potentials. The biomarker potential of cytidine in plasma was reported in two studies [17,22]. The vitreous humor is another biological sample that has been extensively studied to explore novel biomarkers for DR. Haines and Wang confirmed the biomarker potential of pyruvate in the vitreous humor of PDR patients, respectively [38,43]. Predicted biomarkers and their diagnostic performance are detailed below.

3.1. Potential Biomarkers in Human Serum

There are two studies that documented potential biomarkers of DR in human serum [27,30]. Xuan et al. used multiplatform-based metabolomics to generate the metabolic profile of serum samples from 689 subjects with DR and 216 subjects with diabetes without DR [27]. The biomarker panel containing 12-hydroxyeicosatetraenoic acid (12-HETE) and 2-piperidone exhibited good performance for DR diagnosis. The AUC, sensitivity, and specificity of this panel were 0.946, 0.894, and 0.919, respectively, suggesting a potential value as a biomarker for differentiating DR from diabetes. Notably, the biomarker panel also exhibited good performance in differentiating NPDR from diabetes (AUC = 0.958, sensitivity = 0.929, specificity = 0.901). Zuo et al. performed a widely targeted metabolomics based on ultra-performance liquid chromatography-electrospray ionization-tandem mass spectrometry (UPLC-ESI-MS/MS) in the serum samples from 69 subjects with DR and 69 subjects with diabetes without DR [30]. A biomarker model called multidimensional network biomarkers consisting of linoleic acid, nicotinuric acid, ornithine, and phenylacetylglutamine was established. The AUC, sensitivity, and specificity of the MDNBs were 0.92, 0.96, and 0.78, respectively.

3.2. Potential Biomarkers in Human Plasma

Eight studies have reported potential novel biomarkers for DR in human plasma [17,18,19,20,22,24,47,48]. Xia et al. investigated the relationship between pyrimidine metabolites and DR, and identified cytidine as a potential biomarker (AUC = 0.849, sensitivity = 0.737, specificity = 0.919) [17]. Similarly, Xia et al. investigated the relationship between purine metabolites and DR, and identified adenosine as a potential biomarker (AUC = 0.913, sensitivity = 0.947, specificity = 1) [18]. Chen et al. performed metabolomics using GC-MS and found that 1,5-gluconolactone, 2-deoxyribonic acid, gluconic acid, and urea exhibited the potential of a biomarker (AUC = 0.71, 0.68, 0.72, 0.69, respectively) [19]. Rhee et al. recruited 183 patients with type 2 diabetes (52 PDR, 72 NPDR, and 59 NDR) and analyzed their plasma metabolic profiles using ultra-performance liquid chromatography–quadrupole/time-of-flight mass spectrometry (UPLC–Q–TOF–MS) and gas chromatography (GC)–TOF–MS [20]. Their results show that glutamine and glutamic acid were the most differential metabolites and their ratio showed a potential diagnostic value for DR (AUC = 0.742). To identify novel metabolite markers for PDR, Zhu et al. performed metabolomics based on LC-MS in 21 subjects with PDR and 21 subjects with a duration of diabetes of ≥10 years but without DR, and found fumaric acid, uridine, acetic acid, and cytidine to have biomarker potentials (AUC = 0.96, 0.95, 1.0, 0.95, respectively) [22]. Notably, the biomarker potential of cytidine was again demonstrated, which is consistent with the previous study by Xia et al. [17]. Sun et al. recruited 21 patients with PDR, 21 patients with NPDR and 32 patients with type 2 diabetes without DR, and used ultrahigh-performance liquid Q-Exactive mass spectrometry (UPLC-QE-MS) to analyze plasma’s metabolic profile [24]. They established a formula based on the plasma concentration of pseudouridine to calculate the DR risk score: risk score = −0.23 × Ln (pseudouridine) + 1.88. The AUC of the risk score for DR was 0.80, with 97.6% sensitivity and 53.1% specificity. Another formula based on the levels of pseudouridine, N-acetyltryptophan, leucylleucine, and glutamate, was established to calculate the PDR risk score: risk score = 0.23 × Ln(pseudouridine) + 0.16 × Ln(N-acetyltryptophan)-0.065 × Ln(leucylleucine) + 0.11 × Ln(glutamate) − 3.63. The AUC of the risk score for PDR was 0.82, with 76.2% sensitivity and 77.4% specificity. Curovic et al. performed metabolomics and lipidomics analyses to generate the metabolic profile related to DR in 648 individuals with type 1 diabetes [48]. Cox proportional hazard model analysis showed that higher 3,4-dihydroxybutyric acid (3,4-DHBA) was an independent risk marker for DR progression (HR 1.55, 95% CI 1.12–2.15, p = 0.033).

3.3. Potential Biomarkers in Human Vitreous Humor

A total of four studies have investigated potential biomarkers for DR in human vitreous humor [38,40,43,46]. Barba et al. acquired 1H-NMR spectra from vitreous samples of 22 subjects with type 1 diabetes with PDR and 22 non-diabetic subjects, and obtained a model consisting of galactitol and ascorbic acid (AA) that can distinguish PDR and control with 86% sensitivity and 81% specificity [40]. Haines et al. analyzed the vitreous humor of nine patients with PDR and eight non-diabetic patients using UPLC-MS [43]. They performed biomarker analysis using ROC curves, showing that xanthine, proline, citrulline, and pyruvate were the strongest potential predictors of DR (AUC = 1.0, 0.986, 0.972, 0.944 respectively). Wang et al. used gas chromatography coupled with time-of-flight mass spectrometry (GC-TOFMS) to identify potential DR biomarkers in vitreous humor from 28 subjects with type 2 diabetes with PDR and 22 non-diabetic subjects [38]. They found a biomarker panel consisting of pyroglutamic acid and pyruvic acid (AUC = 0.951, sensitivity = 0.955, specificity = 0.857). Zhao et al. performed targeted lipidomics to evaluate oxylipin levels in the vitreous humor using ultra-high-performance liquid-chromatography-multiple reaction monitoring-mass spectrometry/mass spectrometry (UHPLC-MRM-MS/MS) [46]. Vitreous samples were collected from 41 subjects with PDR and 22 non-diabetic subjects. Oxylipins are oxidation products of polyunsaturated fatty acids (PUFAs). According to their results, seven oxylipins were considered as potential biomarkers: docosatetraenoic acid (DTA), eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), arachidonic acid (ARA), ±9(10)-dihydroxy-octadecenoic acid (±9(10)-DiHOME), ±19.20-epoxy-docosapentaenoic acid (±19,20-EpDPE), and ±12(13)- epoxy-octadecenoic acid (±12(13)-EpOME) (AUC = 0.96, 0.803, 0.871, 0.942, 0.805, 0.819, 0.828, respectively).

3.4. Potential Biomarkers in Other Human Samples

DR biomarkers have been predicted in the aqueous humor, cerebrospinal fluid, and feces of humans in three metabolic studies [33,38,47]. Wang et al. also identified potential DR biomarkers in aqueous humor using GC-TOFMS [38]. They recruited 23 subjects with type 2 diabetes with PDR and 25 non-diabetic subjects with cataract, and found a biomarker model consisting of D-2,3-dihydroxypropanoic acid, isocitric acid, fructose 6-phosphate, and L-lactic acid. The AUC of the model was 0.965 with 88% sensitivity and 95.7% specificity. Lin et al. were the first to investigate the metabolomic profile of type 2 diabetes in cerebrospinal fluid [47]. Their study cohort included 19 patients with DR and 14 patients with type 2 diabetes without diabetic microangiopathy. They constructed a multi-marker panel established by alanine, histidine, leucine, pyruvate, tyrosine, and valine showing a high relevance to the occurrence of DR with 0.858 AUC. This multi-marker panel was also validated in plasma with 0.836 AUC. Ye et al. performed 16S rRNA gene sequencing and UPLC-MS-based untargeted metabolomics of fecal samples to investigate the gut metabolome and microbiome [33]. They collected fecal samples from 45 subjects with PDR and 90 subjects with type 2 diabetes without DR. They established a fecal metabolite-based classifier to differentiate PDR and NDR with AUCs of 0.960 and 0.943 in train and test sets. The top 5 differential metabolites in the classifier are alantolactone, desogestrel, adenine, D-erythro-sphinganine, and corosolic acid.

4. Metabolic Pathways Associated with DR

To gain an in-depth understanding of the mechanism underlying metabolic disorders in DR, we counted the differential metabolic pathways reported in plasma and vitreous from DR patients. Purine metabolism, pyrimidine metabolism, arginine and proline metabolism, and glutamate metabolism are the most frequently reported differential pathways in DR metabolomics studies. Details about differential metabolic pathways are summarized in Table 3.

4.1. Pyrimidine Metabolism

Pyrimidine metabolism disorder has been reported in the blood of patients with DR [17,22,23]. Derivatives of pyrimidine exhibit highly potential biological activity as anti-diabetic agents [62,63]. In previous studies [17,22], changed levels of cytidine, a pyrimidine molecule, was observed in patients with DR. Cytidine is the precursor of cytidine triphosphate (CTP), which affects phosphatidylcholine (PC) and phosphatidylethanolamine (PE) biosynthetic pathways. Previous studies reported that phospholipid metabolism is associated with diabetic nephropathy, and that the level of phospholipids decreased with the development of diabetic nephropathy [64]. The mechanism of pyrimidine metabolism in the onset and development of DR still needs further identification and exploration.

4.2. Glutamate Metabolism and Branched-Chain Amino Acid (BCAA) Metabolism

Glutamate metabolism is another affected abnormal metabolic pathway in DR [38]. Glutamate is not only a key signal in the amplification of insulin secretion [65], but is also the major excitatory neurotransmitter in the central nervous system and retina [66,67]. Several studies found increased glutamate and decreased glutamine levels in the vitreous humor of patients with PDR and in diabetic rat retina [68,69,70]. The increased level of glutamate in the retina will cause neurotoxic effects and the activation of ionotropic glutamate receptors in excess, mainly the N-methyl-d-aspartame receptor (NMDAR), resulting in uncontrolled intracellular calcium responses and cell death [71,72,73]. Meanwhile, the levels of leucine, isoleucine, and valine in BCAA metabolism were increased in the serum of DR patients and in the diabetic rat retina [27,74], which are considered to be correlated with the neurotoxic effects of glutamate, which plays an important role in DR neurodegeneration [74]. Therefore, more attention to the abnormal glutamate metabolism and BCAA metabolism may contribute to understanding the pathogenesis of DR.

4.3. Pantothenate and CoA Biosynthesis

Notably, pantothenate and CoA biosynthesis were also altered in both the plasma and vitreous humor of patients with DR [49]. Wang et al. discovered a descending trend of pantothenate in the plasma of PDR patients and an ascending trend of pantothenate in the vitreous [49]. Ma et al. found that the levels of pantothenate and CoA biosynthesis were significantly down-regulated in the urine of patients with diabetic kidney disease [75], which was consistent with Wang’s result in plasma. This phenomenon can probably be explained by a lower pantothenate conversion due to impaired renal tubular reabsorption of vitamins in patients with diabetes complications [76]. A possible explanation for up-regulated levels of pantothenate and CoA biosynthesis in vitreous humor is the mechanism of protecting retinal cells from damage [49]. Endothelial cells were protected from oxidative stress by supplementation with pantothenate [77,78]. Alteration in pantothenate and CoA modulate mitochondrial energy metabolism [79], which is most likely linked to the onset and progression of DR.

4.4. Polyol Pathway

Evidence suggests that the polyol pathway can exacerbate oxidative stress to promote the progression of retinopathy [80]. In the hyperglycemic condition, the polyol pathway of glucose metabolism becomes active in human and rat retinal cells [81,82]. In the polyol pathway, glucose is reduced to sorbitol by aldose reductase (AR), and sorbitol is subsequently oxidized to fructose by sorbitol dehydrogenase (SDH). Fructose can be converted to fructose-3-phosphate by phosphorylation, and then fructose-3-phosphate can be transferred to 3-deoxyglucosone, both of which can be involved in the formation of advanced glycation end products (AGEs) [80]. Reactive oxygen species (ROS) induced by AGEs participate in the oxidative stress process of DR, leading to the impairment of retinal vessels [80]. Secondly, AR can convert NADPH to NADP+ and SDH can convert NAD+ to NADH in the polyol pathway. During the reaction, NADPH is consumed in excess, which results in the reduced synthesis of glutathione (GSH) and the weakened capacity against oxidative stress [83]. In summary, the polyol pathway triggered by hyperglycemia can produce AGEs precursors and expose retinal cells to oxidative stress.

5. Predictions of Metabolism-Based Molecular Targets in DR

To explore potential metabolic enzymes with regulatory potential in DR, we integrated enzymes in DR-related metabolic pathways with a genome-wide association study (GWAS). GWAS refers to multi-center, large sample, and repeatedly verified association studies between genes and diseases at the whole genome level aiming to identify genotype–phenotype associations [84]. Genetic markers (such as SNP) are typed to comprehensively reveal genes related to the onset and development of diseases. Single nucleotide polymorphism (SNP), the most common heritable variation, refers to the polymorphism of the DNA sequence induced by the alteration of a single nucleotide including the conversion or transversion of a single base and the insertion or deletion of bases [85]. GWAS has been successful in identifying risk variants at genetic loci for many diseases including cancers [86,87,88], diabetes [89], and DR [90].
As shown in Figure 2, a total of 23 enzyme-related genes in 6 DR associated metabolic pathways have SNPs through GWAS database analysis. Among these 23 genes, ADCY5, ADCY7, AK5, ENPP3, GUCY1B1, NUDT5, PDE3A, PDE3B, PDE4A, PDE6B, and PGM1 are involved in purine metabolism, DMGDH, PSPH, and SRR are in glycine, serine and threonine metabolism, and ASAH1, CERS6, and GBA2 are in sphingolipid metabolism (Figure 3). As shown in Figure 4, ARG1, CPS1, and NOS1 are involved in arginine biosynthesis, ARG1, AOC1, CKMT1B, and NOS1 are involved in arginine and proline metabolism, and CPS1 and GAD1 are involved in glutamate metabolism.
All 23 genes encode metabolic enzymes in differential metabolic pathways of DR, and there are SNPs associated with diabetes risk in the 23 genes. These indicated a potential link between these enzymes and the pathogenesis of DR. In purine metabolism, PDE3A, PDE3B, PDE4A and PDE6B are closely associated with retinal degeneration under hypoxic or ischemic conditions [91], and the role of ADCY5, ADCY7, AK5, ENPP3, GUCY1B1, NUDT5, and PGM1 in the onset and development of DR is unknown. In addition, AGR1 and NOS1 are involved in arginine metabolism, which was reported to play an important role in the progression of oxidative stress of DR [92,93].

5.1. Arginase 1 and Nitric Oxide Synthase 1

ARG1 encodes arginase 1 catalyzing the hydrolysis of L-arginine to urea and L-ornithine. Ornithine is converted to citrulline, which is converted to arginosuccinate, and finally back to arginine (Figure 4). A recent study showed that high levels of arginase 1 and SNPs (rs2781666 and rs2781665) within the ARG1 are associated with increased type 2 diabetes risk [94]. Elevated concentrations of arginine have been observed in plasma, serum, and the vitreous humor of patients with DR [23,25,26,42]. In addition to the urea cycle, arginine is also involved in the nitric oxide synthesis. Nitric oxide synthase (NOS) catalyzes arginine to citrulline and nitric oxide (NO) [95]. Arginine is the common substrate of nitric oxide synthase and arginase. Under physiological conditions, arginase and NOS compete for the same substrate arginine to produce ornithine and nitric oxide. A study focusing on diabetic cardiomyopathy using H9c2 cells with high glucose treatment found that increased arginase expression results in more arginine flowing to the urea cycle, which reduces the production of NO [96]. In this case, NOS is uncoupled and produces superoxide anions [96]. The superoxide anions react with NO to turn into toxic oxidant peroxynitrite, which is a key indicator of oxidative stress [97]. A previous study has found the increased level of nitrotyrosine, the marker of peroxynitrite, in retinas of the streptozotocin (STZ)-induced diabetic mice [92]. Another study using STZ-induced diabetic mice identified that the inhibition of arginase can reduce nitrotyrosine formation [93]. All of these examples show that ARG1 and NOS1 possess the potential to be regulatory and therapeutic targets for preventing or reversing the oxidative stress of DR.

5.2. Phosphodiesterase

Phosphodiesterase (PDE) catalyzes the hydrolysis of cAMP and cGMP, the second messengers that play important roles through multiple intracellular signaling pathways [98,99]. The PDE family consists of 11 members [100]. SNPs within the PDE3A, PDE3B, PDE4A, and PDE6B genes are associated with increased diabetes risk [101,102,103]. A previous study reported that the accumulation of cGMP through inhibiting PDE prevented hypoxia-induced cell death in porcine retinal explants, which reveals the potential for PDE inhibition to reduce retinal degeneration under hypoxic or ischemic conditions [91].

6. Conclusions

Over the past dozen years, the metabolomics of DR has experienced great growth. Many works have been undertaken so far in the field of DR for biomarker discovery. For example, 12-HETE and 2-piperidone in serum, cytidine and 3,4-DHBA in plasma, and pyruvate in vitreous were all identified to have great potentials to be biomarkers. Moreover, subtle alterations in biological pathways provide insight into the mechanisms. Twenty-three enzymes in DR associated metabolic pathways show potential as targets. Among these 23 enzymes, AGR1 and NOS1 are closely related to arginine metabolism, which was reported to play an important role in the progression of the oxidative stress of DR. PDE is responsible for the hydrolysis of cyclic nucleotides and is closely associated with retinal degeneration under hypoxic or ischemic conditions.
Promising progress in identifying novel biomarkers has been made, yet there are also many limitations. Firstly, most studies focused on the identification of biomarkers for distinguishing between DR and DM. However, few studies further analyzed biomarkers for different stages of DR, especially the early stage (NPDR), which is very important for early diagnosis and prevention. Secondly, a large number of potential biomarkers found in some studies are difficult to be validated in others. Differences in study design, race and region, and clinical characteristics, as well as small sample sizes in some studies, may lead to this issue. Hence, comprehensive research should be conducted to analyze the numerous discriminant metabolites in different kinds of samples for the purpose of identifying biomarkers with real clinical diagnostic values. In terms of target prediction, some of the predicted molecular targets, such as ADCY5, ADCY7, AK5, ENPP3, GUCY1B1, and NUDT5, have no further experimental evidence to be associated with DR. Among them, ADCY5 and ADCY7 are worthy of further exploration. ADCY5 and ADCY7 encode adenylate cyclase 5 and adenylate cyclase 7, respectively [104]. Several previous GWASs demonstrated that SNPs (for example rs11708067 and rs11717195) in ADCY5 may be associated with type 2 diabetes [101,105,106,107]. Hodson et al. reported that ADCY5 mRNA expression in islets was decreased when subjects have risk alleles at rs11708067 [108]. They showed that ADCY5 is essential to couple glucose to insulin secretion by converting glucose signals into cAMP production. Predicted molecular targets provide broader exploration space for DR research.
Metabolomics is demonstrating its power, from biomarker discovery to understanding the mechanisms that underlie DR. This has also been made possible as metabolomics has become more widely integrated with other omics, such as GWAS. The application of metabolomics in DR might also be expanded for judging and monitoring the precise treatment.

Author Contributions

Conceptualization, F.Z. and Y.W.; writing—original draft preparation, Q.J.; writing—review and editing, F.Z. and Y.W. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by Training Program of the Major Research Plan of the National Natural Science Foundation of China (92057106), National Key R&D Program of China (2021YFF0702100 and 2021YFA0805100), Natural Science Foundation of China (32171177), Natural Science Foundation of Shanghai (19ZR1440500), Shanghai Jiaotong University-Gaofeng Clinical Medicine Grant Support, and Shanghai Pujiang Program (2019PJD046).

Conflicts of Interest

The authors declare that they have no conflict of interest.

References

  1. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2011, 34 (Suppl. S1), S62–S69. [Google Scholar] [CrossRef] [PubMed]
  2. Cheung, N.; Mitchell, P.; Wong, T.Y. Diabetic retinopathy. Lancet 2010, 376, 124–136. [Google Scholar] [CrossRef]
  3. Wong, T.Y.; Cheung, C.M.; Larsen, M.; Sharma, S.; Simó, R. Diabetic retinopathy. Nat. Rev. Dis. Prim. 2016, 2, 16012. [Google Scholar] [CrossRef] [PubMed]
  4. Leasher, J.L.; Bourne, R.R.A.; Flaxman, S.R.; Jonas, J.B.; Keeffe, J.; Naidoo, K.; Pesudovs, K.; Price, H.; White, R.A.; Wong, T.Y.; et al. Global Estimates on the Number of People Blind or Visually Impaired by Diabetic Retinopathy: A Meta-analysis from 1990 to 2010. Diabetes Care 2016, 39, 1643–1649. [Google Scholar] [CrossRef]
  5. Teo, Z.L.; Tham, Y.C.; Yu, M.; Chee, M.L.; Rim, T.H.; Cheung, N.; Bikbov, M.M.; Wang, Y.X.; Tang, Y.; Lu, Y.; et al. Global Prevalence of Diabetic Retinopathy and Projection of Burden through 2045: Systematic Review and Meta-analysis. Ophthalmology 2021, 128, 1580–1591. [Google Scholar] [CrossRef]
  6. Ting, D.S.W.; Tan, K.-A.; Phua, V.; Tan, G.S.W.; Wong, C.W.; Wong, T.Y. Biomarkers of diabetic retinopathy. Curr. Diabetes Rep. 2016, 16, 125. [Google Scholar] [CrossRef]
  7. Hirsch, I.B.; Brownlee, M. Beyond hemoglobin A1c—Need for additional markers of risk for diabetic microvascular complications. JAMA 2010, 303, 2291–2292. [Google Scholar] [CrossRef]
  8. The Diabetes Control and Complications Trial Research Group. The Effect of Intensive Treatment of Diabetes on the Development and Progression of Long-Term Complications in Insulin-Dependent Diabetes Mellitus. N. Engl. J. Med. 1993, 329, 977–986. [Google Scholar] [CrossRef]
  9. Stitt, A.W.; Curtis, T.M.; Chen, M.; Medina, R.J.; McKay, G.J.; Jenkins, A.; Gardiner, T.A.; Lyons, T.J.; Hammes, H.-P.; Simó, R.; et al. The progress in understanding and treatment of diabetic retinopathy. Prog. Retin. Eye Res. 2016, 51, 156–186. [Google Scholar] [CrossRef]
  10. Ajlan, R.S.; Silva, P.S.; Sun, J.K. Vascular endothelial growth factor and diabetic retinal disease. Semin. Ophthalmol. 2016, 31, 40–48. [Google Scholar] [CrossRef]
  11. Li, A.S.; Veerappan, M.; Mittal, V.; Do, D.V. Anti-VEGF agents in the management of diabetic macular edema. Expert Rev. Ophthalmol. 2020, 15, 285–296. [Google Scholar] [CrossRef]
  12. Bandello, F.; Cicinelli, M.V. 19th EURETINA Congress Keynote Lecture: Diabetic Retinopathy Today. Ophthalmologica 2020, 243, 163–171. [Google Scholar] [CrossRef]
  13. Hammes, H.-P. Diabetic retinopathy: Hyperglycaemia, oxidative stress and beyond. Diabetologia 2018, 61, 29–38. [Google Scholar] [CrossRef]
  14. Clish, C.B. Metabolomics: An emerging but powerful tool for precision medicine. Cold Spring Harb. Mol. Case Study 2015, 1, a000588. [Google Scholar] [CrossRef]
  15. Laíns, I.; Gantner, M.; Murinello, S.; Lasky-Su, J.A.; Miller, J.W.; Friedlander, M.; Husain, D. Metabolomics in the study of retinal health and disease. Prog. Retin. Eye Res. 2019, 69, 57–79. [Google Scholar] [CrossRef]
  16. Li, X.; Luo, X.; Lu, X.; Duan, J.; Xu, G. Metabolomics study of diabetic retinopathy using gas chromatography-mass spectrometry: A comparison of stages and subtypes diagnosed by Western and Chinese medicine. Mol. BioSyst. 2011, 7, 2228–2237. [Google Scholar] [CrossRef]
  17. Xia, J.F.; Wang, Z.H.; Liang, Q.L.; Wang, Y.M.; Li, P.; Luo, G.A. Correlations of six related pyrimidine metabolites and diabetic retinopathy in Chinese type 2 diabetic patients. Clin. Chim. Acta 2011, 412, 940–945. [Google Scholar] [CrossRef]
  18. Xia, J.; Wang, Z.; Zhang, F. Association between Related Purine Metabolites and Diabetic Retinopathy in Type 2 Diabetic Patients. Int. J. Endocrinol. 2014, 2014, 651050. [Google Scholar] [CrossRef]
  19. Chen, L.; Cheng, C.Y.; Choi, H.; Ikram, M.K.; Sabanayagam, C.; Tan, G.S.; Tian, D.; Zhang, L.; Venkatesan, G.; Tai, E.S.; et al. Plasma Metabonomic Profiling of Diabetic Retinopathy. Diabetes 2016, 65, 1099–1108. [Google Scholar] [CrossRef]
  20. Rhee, S.Y.; Jung, E.S.; Park, H.M.; Jeong, S.J.; Kim, K.; Chon, S.; Yu, S.Y.; Woo, J.T.; Lee, C.H. Plasma glutamine and glutamic acid are potential biomarkers for predicting diabetic retinopathy. Metabolomics 2018, 14, 89. [Google Scholar] [CrossRef] [Green Version]
  21. Peng, L.; Sun, B.; Liu, M.; Huang, J.; Liu, Y.; Xie, Z.; He, J.; Chen, L.; Wang, D.; Zhu, Y.; et al. Plasma metabolic profile reveals PGF2α protecting against non-proliferative diabetic retinopathy in patients with type 2 diabetes. Biochem. Biophys. Res. Commun. 2018, 496, 1276–1283. [Google Scholar] [CrossRef]
  22. Zhu, X.R.; Yang, F.Y.; Lu, J.; Zhang, H.R.; Sun, R.; Zhou, J.B.; Yang, J.K. Plasma metabolomic profiling of proliferative diabetic retinopathy. Nutr. Metab. 2019, 16, 37. [Google Scholar] [CrossRef]
  23. Sumarriva, K.; Uppal, K.; Ma, C.; Herren, D.J.; Wang, Y.; Chocron, I.M.; Warden, C.; Mitchell, S.L.; Burgess, L.G.; Goodale, M.P.; et al. Arginine and Carnitine Metabolites Are Altered in Diabetic Retinopathy. Investig. Ophthalmol. Vis. Sci. 2019, 60, 3119–3126. [Google Scholar] [CrossRef]
  24. Sun, Y.; Zou, H.; Li, X.; Xu, S.; Liu, C. Plasma Metabolomics Reveals Metabolic Profiling for Diabetic Retinopathy and Disease Progression. Front. Endocrinol. 2021, 12, 757088. [Google Scholar] [CrossRef] [PubMed]
  25. Peters, K.S.; Rivera, E.; Warden, C.; Harlow, P.A.; Mitchell, S.L.; Calcutt, M.W.; Samuels, D.C.; Brantley, M.A., Jr. Plasma Arginine and Citrulline are Elevated in Diabetic Retinopathy. Am. J. Ophthalmol. 2021, 235, 154–162. [Google Scholar] [CrossRef]
  26. Abhary, S.; Kasmeridis, N.; Burdon, K.P.; Kuot, A.; Whiting, M.J.; Yew, W.P.; Petrovsky, N.; Craig, J.E. Diabetic retinopathy is associated with elevated serum asymmetric and symmetric dimethylarginines. Diabetes Care 2009, 32, 2084–2086. [Google Scholar] [CrossRef]
  27. Xuan, Q.; Ouyang, Y.; Wang, Y.; Wu, L.; Li, H.; Luo, Y.; Zhao, X.; Feng, D.; Qin, W.; Hu, C.; et al. Multiplatform Metabolomics Reveals Novel Serum Metabolite Biomarkers in Diabetic Retinopathy Subjects. Adv. Sci. 2020, 7, 2001714. [Google Scholar] [CrossRef] [PubMed]
  28. Xuan, Q.; Zheng, F.; Yu, D.; Ouyang, Y.; Zhao, X.; Hu, C.; Xu, G. Rapid lipidomic profiling based on ultra-high performance liquid chromatography-mass spectrometry and its application in diabetic retinopathy. Anal. Bioanal. Chem. 2020, 412, 3585–3594. [Google Scholar] [CrossRef]
  29. Yun, J.H.; Kim, J.M.; Jeon, H.J.; Oh, T.; Choi, H.J.; Kim, B.J. Metabolomics profiles associated with diabetic retinopathy in type 2 diabetes patients. PLoS ONE 2020, 15, e0241365. [Google Scholar] [CrossRef]
  30. Zuo, J.; Lan, Y.; Hu, H.; Hou, X.; Li, J.; Wang, T.; Zhang, H.; Zhang, N.; Guo, C.; Peng, F.; et al. Metabolomics-based multidimensional network biomarkers for diabetic retinopathy identification in patients with type 2 diabetes mellitus. BMJ Open Diabetes Res. Care 2021, 9, e001443. [Google Scholar] [CrossRef] [PubMed]
  31. Yousri, N.A.; Suhre, K.; Yassin, E.; Al-Shakaki, A.; Robay, A.; Elshafei, M.; Chidiac, O.; Hunt, S.C.; Crystal, R.G.; Fakhro, K.A. Metabolic and Metabo-Clinical Signatures of Type 2 Diabetes, Obesity, Retinopathy, and Dyslipidemia. Diabetes 2021, 71, 184–205. [Google Scholar] [CrossRef] [PubMed]
  32. Koehrer, P.; Saab, S.; Berdeaux, O.; Isaïco, R.; Grégoire, S.; Cabaret, S.; Bron, A.M.; Creuzot-Garcher, C.P.; Bretillon, L.; Acar, N. Erythrocyte phospholipid and polyunsaturated fatty acid composition in diabetic retinopathy. PLoS ONE 2014, 9, e106912. [Google Scholar] [CrossRef] [PubMed]
  33. Ye, P.; Zhang, X.; Xu, Y.; Xu, J.; Song, X.; Yao, K. Alterations of the Gut Microbiome and Metabolome in Patients with Proliferative Diabetic Retinopathy. Front. Microbiol. 2021, 12, 667632. [Google Scholar] [CrossRef] [PubMed]
  34. Zhou, Z.; Zheng, Z.; Xiong, X.; Chen, X.; Peng, J.; Yao, H.; Pu, J.; Chen, Q.; Zheng, M. Gut Microbiota Composition and Fecal Metabolic Profiling in Patients with Diabetic Retinopathy. Front. Cell Dev. Biol. 2021, 9, 732204. [Google Scholar] [CrossRef]
  35. Fort, P.E.; Rajendiran, T.M.; Soni, T.; Byun, J.; Shan, Y.; Looker, H.C.; Nelson, R.G.; Kretzler, M.; Michailidis, G.; Roger, J.E.; et al. Diminished retinal complex lipid synthesis and impaired fatty acid β-oxidation associated with human diabetic retinopathy. JCI Insight 2021, 6, e152109. [Google Scholar] [CrossRef]
  36. Jin, H.; Zhu, B.; Liu, X.; Jin, J.; Zou, H. Metabolic characterization of diabetic retinopathy: An (1)H-NMR-based metabolomic approach using human aqueous humor. J. Pharm. Biomed. Anal. 2019, 174, 414–421. [Google Scholar] [CrossRef]
  37. Kunikata, H.; Ida, T.; Sato, K.; Aizawa, N.; Sawa, T.; Tawarayama, H.; Murayama, N.; Fujii, S.; Akaike, T.; Nakazawa, T. Metabolomic profiling of reactive persulfides and polysulfides in the aqueous and vitreous humors. Sci. Rep. 2017, 7, 41984. [Google Scholar] [CrossRef]
  38. Wang, H.; Fang, J.; Chen, F.; Sun, Q.; Xu, X.; Lin, S.H.; Liu, K. Metabolomic profile of diabetic retinopathy: A GC-TOFMS-based approach using vitreous and aqueous humor. Acta Diabetol. 2020, 57, 41–51. [Google Scholar] [CrossRef]
  39. Young, S.P.; Nessim, M.; Falciani, F.; Trevino, V.; Banerjee, S.P.; Scott, R.A.H.; Murray, P.I.; Wallace, G.R. Metabolomic analysis of human vitreous humor differentiates ocular inflammatory disease. Mol. Vis. 2009, 15, 1210–1217. [Google Scholar]
  40. Barba, I.; Garcia-Ramírez, M.; Hernández, C.; Alonso, M.A.; Masmiquel, L.; García-Dorado, D.; Simó, R. Metabolic fingerprints of proliferative diabetic retinopathy: An 1H-NMR-based metabonomic approach using vitreous humor. Investig. Ophthalmol. Vis. Sci. 2010, 51, 4416–4421. [Google Scholar] [CrossRef]
  41. Schwartzman, M.L.; Iserovich, P.; Gotlinger, K.; Bellner, L.; Dunn, M.W.; Sartore, M.; Grazia Pertile, M.; Leonardi, A.; Sathe, S.; Beaton, A.; et al. Profile of lipid and protein autacoids in diabetic vitreous correlates with the progression of diabetic retinopathy. Diabetes 2010, 59, 1780–1788. [Google Scholar] [CrossRef] [Green Version]
  42. Paris, L.P.; Johnson, C.H.; Aguilar, E.; Usui, Y.; Cho, K.; Hoang, L.T.; Feitelberg, D.; Benton, H.P.; Westenskow, P.D.; Kurihara, T.; et al. Global metabolomics reveals metabolic dysregulation in ischemic retinopathy. Metabolomics 2016, 12, 15. [Google Scholar] [CrossRef] [PubMed]
  43. Haines, N.R.; Manoharan, N.; Olson, J.L.; D’Alessandro, A.; Reisz, J.A. Metabolomics Analysis of Human Vitreous in Diabetic Retinopathy and Rhegmatogenous Retinal Detachment. J. Proteome Res. 2018, 17, 2421–2427. [Google Scholar] [CrossRef] [PubMed]
  44. Lin, A.L.; Roman, R.J.; Regan, K.A.; Bolch, C.A.; Chen, C.J.; Iyer, S.S.R. Eicosanoid Profiles in the Vitreous Humor of Patients with Proliferative Diabetic Retinopathy. Int. J. Mol. Sci. 2020, 21, 7451. [Google Scholar] [CrossRef] [PubMed]
  45. Tomita, Y.; Cagnone, G.; Fu, Z.; Cakir, B.; Kotoda, Y.; Asakage, M.; Wakabayashi, Y.; Hellström, A.; Joyal, J.S.; Talukdar, S.; et al. Vitreous metabolomics profiling of proliferative diabetic retinopathy. Diabetologia 2021, 64, 70–82. [Google Scholar] [CrossRef] [PubMed]
  46. Zhao, T.; Wang, Y.; Guo, X.; Li, H.; Jiang, W.; Xiao, Y.; Deng, B.; Sun, Y. Altered oxylipin levels in human vitreous indicate imbalance in pro-/anti-inflammatory homeostasis in proliferative diabetic retinopathy. Exp. Eye Res. 2022, 214, 108799. [Google Scholar] [CrossRef] [PubMed]
  47. Lin, H.T.; Cheng, M.L.; Lo, C.J.; Lin, G.; Lin, S.F.; Yeh, J.T.; Ho, H.Y.; Lin, J.R.; Liu, F.C. (1)H Nuclear Magnetic Resonance (NMR)-Based Cerebrospinal Fluid and Plasma Metabolomic Analysis in Type 2 Diabetic Patients and Risk Prediction for Diabetic Microangiopathy. J. Clin. Med. 2019, 8, 874. [Google Scholar] [CrossRef]
  48. Curovic, V.R.; Suvitaival, T.; Mattila, I.; Ahonen, L.; Trošt, K.; Theilade, S.; Hansen, T.W.; Legido-Quigley, C.; Rossing, P. Circulating Metabolites and Lipids Are Associated to Diabetic Retinopathy in Individuals with Type 1 Diabetes. Diabetes 2020, 69, 2217–2226. [Google Scholar] [CrossRef]
  49. Wang, H.; Li, S.; Wang, C.; Wang, Y.; Fang, J.; Liu, K. Plasma and Vitreous Metabolomics Profiling of Proliferative Diabetic Retinopathy. Investig. Ophthalmol. Vis. Sci. 2022, 63, 17. [Google Scholar] [CrossRef]
  50. Quek, D.Q.Y.; He, F.; Sultana, R.; Banu, R.; Chee, M.L.; Nusinovici, S.; Thakur, S.; Qian, C.; Cheng, C.-Y.; Wong, T.Y.; et al. Novel Serum and Urinary Metabolites Associated with Diabetic Retinopathy in Three Asian Cohorts. Metabolites 2021, 11, 614. [Google Scholar] [CrossRef]
  51. Wang, X.; Li, Y.; Xie, M.; Deng, L.; Zhang, M.; Xie, X. Urine metabolomics study of Bushen Huoxue Prescription on diabetic retinopathy rats by UPLC-Q-exactive Orbitrap-MS. Biomed. Chromatogr. 2020, 34, e4792. [Google Scholar] [CrossRef]
  52. Sas, K.M.; Lin, J.; Rajendiran, T.M.; Soni, T.; Nair, V.; Hinder, L.M.; Jagadish, H.V.; Gardner, T.W.; Abcouwer, S.F.; Brosius, F.C., III; et al. Shared and distinct lipid-lipid interactions in plasma and affected tissues in a diabetic mouse model. J. Lipid Res. 2018, 59, 173–183. [Google Scholar] [CrossRef]
  53. Ge, N.; Kong, L.; Zhang, A.-H.; Sun, Y.; Zhao, M.-Q.; Zhang, B.; Xu, L.; Ke, X.; Sun, H.; Wang, X.-J. Identification of key lipid metabolites during metabolic dysregulation in the diabetic retinopathy disease mouse model and efficacy of Keluoxin capsule using an UHPLC-MS-based non-targeted lipidomics approach. RSC Adv. 2021, 11, 5491–5505. [Google Scholar] [CrossRef]
  54. Wiggenhauser, L.M.; Qi, H.; Stoll, S.J.; Metzger, L.; Bennewitz, K.; Poschet, G.; Krenning, G.; Hillebrands, J.L.; Hammes, H.P.; Kroll, J. Activation of Retinal Angiogenesis in Hyperglycemic pdx1 (-/-) Zebrafish Mutants. Diabetes 2020, 69, 1020–1031. [Google Scholar] [CrossRef]
  55. Chetwynd, A.J.; Dunn, W.B.; Rodriguez-Blanco, G. Collection and Preparation of Clinical Samples for Metabolomics. Adv. Exp. Med. Biol. 2017, 965, 19–44. [Google Scholar] [CrossRef]
  56. Vignoli, A.; Tenori, L.; Morsiani, C.; Turano, P.; Capri, M.; Luchinat, C. Serum or Plasma (and Which Plasma), That Is the Question. J. Proteome Res. 2022, 21, 1061–1072. [Google Scholar] [CrossRef]
  57. Yu, Z.; Kastenmüller, G.; He, Y.; Belcredi, P.; Möller, G.; Prehn, C.; Mendes, J.; Wahl, S.; Roemisch-Margl, W.; Ceglarek, U.; et al. Differences between Human Plasma and Serum Metabolite Profiles. PLoS ONE 2011, 6, e21230. [Google Scholar] [CrossRef]
  58. Floyd, J.L.; Grant, M.B. The gut–eye axis: Lessons learned from murine models. Ophthalmol. Ther. 2020, 9, 499–513. [Google Scholar] [CrossRef]
  59. Emwas, A.-H.M. The strengths and weaknesses of NMR spectroscopy and mass spectrometry with particular focus on metabolomics research. In Metabonomics; Humana Press: New York, NY, USA, 2015; pp. 161–193. [Google Scholar]
  60. Barnes, S.; Benton, H.P.; Casazza, K.; Cooper, S.J.; Cui, X.; Du, X.; Engler, J.; Kabarowski, J.H.; Li, S.; Pathmasiri, W. Training in metabolomics research. I. Designing the experiment, collecting and extracting samples and generating metabolomics data. J. Mass Spectrom. 2016, 51, 461–475. [Google Scholar] [CrossRef]
  61. Johnson, C.H.; Ivanisevic, J.; Siuzdak, G. Metabolomics: Beyond biomarkers and towards mechanisms. Nat. Rev. Mol. Cell Biol. 2016, 17, 451–459. [Google Scholar] [CrossRef]
  62. Fang, Y.; Xu, J.; Li, Z.; Yang, Z.; Xiong, L.; Jin, Y.; Wang, Q.; Xie, S.; Zhu, W.; Chang, S. Design and synthesis of novel pyrimido [5, 4-d] pyrimidine derivatives as GPR119 agonist for treatment of type 2 diabetes. Bioorganic Med. Chem. 2018, 26, 4080–4087. [Google Scholar] [CrossRef]
  63. Balestri, F.; Quattrini, L.; Coviello, V.; Sartini, S.; Da Settimo, F.; Cappiello, M.; Moschini, R.; Del Corso, A.; Mura, U.; La Motta, C. Acid derivatives of pyrazolo [1, 5-a] pyrimidine as aldose reductase differential inhibitors. Cell Chem. Biol. 2018, 25, 1414–1418.e3. [Google Scholar] [CrossRef]
  64. Pang, L.-Q.; Liang, Q.-L.; Wang, Y.-M.; Ping, L.; Luo, G.-A. Simultaneous determination and quantification of seven major phospholipid classes in human blood using normal-phase liquid chromatography coupled with electrospray mass spectrometry and the application in diabetes nephropathy. J. Chromatogr. B 2008, 869, 118–125. [Google Scholar] [CrossRef] [PubMed]
  65. Gheni, G.; Ogura, M.; Iwasaki, M.; Yokoi, N.; Minami, K.; Nakayama, Y.; Harada, K.; Hastoy, B.; Wu, X.; Takahashi, H.; et al. Glutamate Acts as a Key Signal Linking Glucose Metabolism to Incretin/cAMP Action to Amplify Insulin Secretion. Cell Rep. 2014, 9, 661–673. [Google Scholar] [CrossRef]
  66. Bringmann, A.; Pannicke, T.; Biedermann, B.; Francke, M.; Iandiev, I.; Grosche, J.; Wiedemann, P.; Albrecht, J.; Reichenbach, A. Role of retinal glial cells in neurotransmitter uptake and metabolism. Neurochem. Int. 2009, 54, 143–160. [Google Scholar] [CrossRef]
  67. Brosnan, J.T.; Brosnan, M.E. Glutamate: A truly functional amino acid. Amino Acids 2013, 45, 413–418. [Google Scholar] [CrossRef]
  68. Ambati, J.; Chalam, K.V.; Chawla, D.K.; D’Angio, C.T.; Guillet, E.G.; Rose, S.J.; Vanderlinde, R.E.; Ambati, B.K. Elevated γ-Aminobutyric Acid, Glutamate, and Vascular Endothelial Growth Factor Levels in the Vitreous of Patients with Proliferative Diabetic Retinopathy. Arch. Ophthalmol. 1997, 115, 1161–1166. [Google Scholar] [CrossRef]
  69. Ishikawa, S.; Nakazawa, M.; Ishikawa, A.; Ishiguro, S.-I.; Tamai, M. Alteration of glutamine concentration in the vitreous humor in patients with proliferative vitreoretinopathy. Curr. Eye Res. 1995, 14, 191–197. [Google Scholar] [CrossRef]
  70. Lieth, E.; Barber, A.J.; Xu, B.; Dice, C.; Ratz, M.J.; Tanase, D.; Strother, J.M. Glial reactivity and impaired glutamate metabolism in short-term experimental diabetic retinopathy. Penn State Retina Research Group. Diabetes 1998, 47, 815–820. [Google Scholar] [CrossRef]
  71. Bogdanov, P.; Corraliza, L.; Villena, J.A.; Carvalho, A.R.; Garcia-Arumí, J.; Ramos, D.; Ruberte, J.; Simó, R.; Hernández, C. The db/db Mouse: A Useful Model for the Study of Diabetic Retinal Neurodegeneration. PLoS ONE 2014, 9, e97302. [Google Scholar] [CrossRef]
  72. Ng, Y.-K.; Zeng, X.-X.; Ling, E.-A. Expression of glutamate receptors and calcium-binding proteins in the retina of streptozotocin-induced diabetic rats. Brain Res. 2004, 1018, 66–72. [Google Scholar] [CrossRef] [PubMed]
  73. Santiago, A.R.; Gaspar, J.M.; Baptista, F.I.; Cristóvão, A.J.; Santos, P.F.; Kamphuis, W.; Ambrósio, A.F. Diabetes changes the levels of ionotropic glutamate receptors in the rat retina. Mol. Vis. 2009, 15, 1620–1630. [Google Scholar]
  74. Ola, M.S.; Alhomida, A.S.; LaNoue, K.F. Gabapentin Attenuates Oxidative Stress and Apoptosis in the Diabetic Rat Retina. Neurotox. Res. 2019, 36, 81–90. [Google Scholar] [CrossRef] [PubMed]
  75. Ma, T.; Liu, T.; Xie, P.; Jiang, S.; Yi, W.; Dai, P.; Guo, X. UPLC-MS-based urine nontargeted metabolic profiling identifies dysregulation of pantothenate and CoA biosynthesis pathway in diabetic kidney disease. Life Sci. 2020, 258, 118160. [Google Scholar] [CrossRef]
  76. Hatano, M.; Hodges, R.E.; Evans, T.C.; Hagemann, R.F.; Leeper, D.B.; Bean, W.B.; Krehl, W.A. Urinary excretion of pantothenic acid by diabetic patients and by alloxan-diabetic rats. Am. J. Clin. Nutr. 1967, 20, 960–967. [Google Scholar] [CrossRef]
  77. Slyshenkov, V.S.; Rakowska, M.; Wojtczak, L. Protective effect of pantothenic acid and related compounds against permeabilization of Ehrlich ascites tumour cells by digitonin. Acta Biochim. Pol. 1996, 43, 407–410. [Google Scholar] [CrossRef]
  78. Demirci, B.; Demir, O.; Dost, T.; Birincioglu, M. Protective effect of vitamin B5 (dexpanthenol) on cardiovascular damage induced by streptozocin in rats. Bratisl. Lek. Listy 2014, 115, 190–196. [Google Scholar] [CrossRef]
  79. Depeint, F.; Bruce, W.R.; Shangari, N.; Mehta, R.; O’Brien, P.J. Mitochondrial function and toxicity: Role of B vitamins on the one-carbon transfer pathways. Chem. Biol. Interact. 2006, 163, 113–132. [Google Scholar] [CrossRef]
  80. Kang, Q.; Yang, C. Oxidative stress and diabetic retinopathy: Molecular mechanisms, pathogenetic role and therapeutic implications. Redox Biol. 2020, 37, 101799. [Google Scholar] [CrossRef]
  81. Oates, P.J. Polyol Pathway and Diabetic Peripheral Neuropathy. In International Review of Neurobiology; Academic Press: Cambridge, MA, USA, 2002; Volume 50, pp. 325–392. [Google Scholar]
  82. Dagher, Z.; Park, Y.S.; Asnaghi, V.; Hoehn, T.; Gerhardinger, C.; Lorenzi, M. Studies of Rat and Human Retinas Predict a Role for the Polyol Pathway in Human Diabetic Retinopathy. Diabetes 2004, 53, 2404–2411. [Google Scholar] [CrossRef]
  83. Li, C.; Miao, X.; Li, F.; Wang, S.; Liu, Q.; Wang, Y.; Sun, J. Oxidative Stress-Related Mechanisms and Antioxidant Therapy in Diabetic Retinopathy. Oxidative Med. Cell. Longev. 2017, 2017, 9702820. [Google Scholar] [CrossRef] [PubMed]
  84. Tam, V.; Patel, N.; Turcotte, M.; Bossé, Y.; Paré, G.; Meyre, D. Benefits and limitations of genome-wide association studies. Nat. Rev. Genet. 2019, 20, 467–484. [Google Scholar] [CrossRef]
  85. Shastry, B.S. SNPs: Impact on gene function and phenotype. Methods Mol. Biol. 2009, 578, 3–22. [Google Scholar] [CrossRef]
  86. Visscher, P.M.; Brown, M.A.; McCarthy, M.I.; Yang, J. Five years of GWAS discovery. Am. J. Hum. Genet. 2012, 90, 7–24. [Google Scholar] [CrossRef]
  87. Milne, R.L.; Kuchenbaecker, K.B.; Michailidou, K.; Beesley, J.; Kar, S.; Lindström, S.; Hui, S.; Lemaçon, A.; Soucy, P.; Dennis, J.; et al. Identification of ten variants associated with risk of estrogen-receptor-negative breast cancer. Nat. Genet. 2017, 49, 1767–1778. [Google Scholar] [CrossRef]
  88. Sud, A.; Kinnersley, B.; Houlston, R.S. Genome-wide association studies of cancer: Current insights and future perspectives. Nat. Rev. Cancer 2017, 17, 692–704. [Google Scholar] [CrossRef]
  89. Zhao, W.; Rasheed, A.; Tikkanen, E.; Lee, J.J.; Butterworth, A.S.; Howson, J.M.M.; Assimes, T.L.; Chowdhury, R.; Orho-Melander, M.; Damrauer, S.; et al. Identification of new susceptibility loci for type 2 diabetes and shared etiological pathways with coronary heart disease. Nat. Genet. 2017, 49, 1450–1457. [Google Scholar] [CrossRef]
  90. Khan, N.; Paterson, A.D.; Roshandel, D.; Raza, A.; Ajmal, M.; Waheed, N.K.; Azam, M.; Qamar, R. Association of IGF1 and VEGFA polymorphisms with diabetic retinopathy in Pakistani population. Acta Diabetol. 2020, 57, 237–245. [Google Scholar] [CrossRef]
  91. Olivares-González, L.; Martínez-Fernández de la Cámara, C.; Hervás, D.; Marín, M.P.; Lahoz, A.; Millán, J.M.; Rodrigo, R. cGMP-Phosphodiesterase Inhibition Prevents Hypoxia-Induced Cell Death Activation in Porcine Retinal Explants. PLoS ONE 2016, 11, e0166717. [Google Scholar] [CrossRef]
  92. Patel, C.; Rojas, M.; Narayanan, S.P.; Zhang, W.; Xu, Z.; Lemtalsi, T.; Jittiporn, K.; Caldwell, R.; Caldwell, R. Arginase as a Mediator of Diabetic Retinopathy. Front. Immunol. 2013, 4, 173. [Google Scholar] [CrossRef]
  93. Elms, S.C.; Toque, H.A.; Rojas, M.; Xu, Z.; Caldwell, R.W.; Caldwell, R.B. The role of arginase I in diabetes-induced retinal vascular dysfunction in mouse and rat models of diabetes. Diabetologia 2013, 56, 654–662. [Google Scholar] [CrossRef] [PubMed]
  94. Fawad Ali Shah, S.; Iqbal, T.; Naveed, N.; Akram, S.; Arshad Rafiq, M.; Hussain, S. ARG1 single nucleotide polymorphisms rs2781666 and rs2781665 confer risk of Type 2 diabetes mellitus. EXCLI J. 2018, 17, 847–855. [Google Scholar] [CrossRef]
  95. Wu, G.; Morris, S.M., Jr. Arginine metabolism: Nitric oxide and beyond. Biochem. J. 1998, 336, 1–17. [Google Scholar] [CrossRef] [PubMed]
  96. Zhou, L.; Sun, C.-B.; Liu, C.; Fan, Y.; Zhu, H.-Y.; Wu, X.-W.; Hu, L.; Li, Q.-P. Upregulation of arginase activity contributes to intracellular ROS production induced by high glucose in H9c2 cells. Int. J. Clin. Exp. Pathol. 2015, 8, 2728–2736. [Google Scholar]
  97. Shosha, E.; Fouda, A.Y.; Narayanan, S.P.; Caldwell, R.W.; Caldwell, R.B. Is the Arginase Pathway a Novel Therapeutic Avenue for Diabetic Retinopathy? J. Clin. Med. 2020, 9, 425. [Google Scholar] [CrossRef]
  98. Manganiello, V.C.; Fontana, J.; Degerman, E.; Ahmad, F. Cyclic Nucleotide Phosphodiesterases. In Encyclopedia of Respiratory Medicine; Laurent, G.J., Shapiro, S.D., Eds.; Academic Press: Oxford, UK, 2006; pp. 589–594. [Google Scholar]
  99. Kilanowska, A.; Ziółkowska, A. Role of Phosphodiesterase in the Biology and Pathology of Diabetes. Int. J. Mol. Sci. 2020, 21, 8244. [Google Scholar] [CrossRef]
  100. Heimann, E.; Jones, H.A.; Resjö, S.; Manganiello, V.C.; Stenson, L.; Degerman, E. Expression and Regulation of Cyclic Nucleotide Phosphodiesterases in Human and Rat Pancreatic Islets. PLoS ONE 2010, 5, e14191. [Google Scholar] [CrossRef]
  101. Vujkovic, M.; Keaton, J.M.; Lynch, J.A.; Miller, D.R.; Zhou, J.; Tcheandjieu, C.; Huffman, J.E.; Assimes, T.L.; Lorenz, K.; Zhu, X.; et al. Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1.4 million participants in a multi-ancestry meta-analysis. Nat. Genet. 2020, 52, 680–691. [Google Scholar] [CrossRef]
  102. Mahajan, A.; Taliun, D.; Thurner, M.; Robertson, N.R.; Torres, J.M.; Rayner, N.W.; Payne, A.J.; Steinthorsdottir, V.; Scott, R.A.; Grarup, N.; et al. Fine-mapping type 2 diabetes loci to single-variant resolution using high-density imputation and islet-specific epigenome maps. Nat. Genet. 2018, 50, 1505–1513. [Google Scholar] [CrossRef]
  103. Li, Y.R.; Li, J.; Zhao, S.D.; Bradfield, J.P.; Mentch, F.D.; Maggadottir, S.M.; Hou, C.; Abrams, D.J.; Chang, D.; Gao, F.; et al. Meta-analysis of shared genetic architecture across ten pediatric autoimmune diseases. Nat. Med. 2015, 21, 1018–1027. [Google Scholar] [CrossRef]
  104. Defer, N.; Best-Belpomme, M.; Hanoune, J. Tissue specificity and physiological relevance of various isoforms of adenylyl cyclase. Am. J. Physiol. Ren. Physiol. 2000, 279, F400–F416. [Google Scholar] [CrossRef] [PubMed]
  105. Mahajan, A.; Go, M.J.; Zhang, W.; Below, J.E.; Gaulton, K.J.; Ferreira, T.; Horikoshi, M.; Johnson, A.D.; Ng, M.C.; Prokopenko, I.; et al. Genome-wide trans-ancestry meta-analysis provides insight into the genetic architecture of type 2 diabetes susceptibility. Nat. Genet. 2014, 46, 234–244. [Google Scholar] [CrossRef] [PubMed]
  106. Dupuis, J.; Langenberg, C.; Prokopenko, I.; Saxena, R.; Soranzo, N.; Jackson, A.U.; Wheeler, E.; Glazer, N.L.; Bouatia-Naji, N.; Gloyn, A.L.; et al. New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. Nat. Genet. 2010, 42, 105–116. [Google Scholar] [CrossRef] [PubMed]
  107. Xue, A.; Wu, Y.; Zhu, Z.; Zhang, F.; Kemper, K.E.; Zheng, Z.; Yengo, L.; Lloyd-Jones, L.R.; Sidorenko, J.; Wu, Y.; et al. Genome-wide association analyses identify 143 risk variants and putative regulatory mechanisms for type 2 diabetes. Nat. Commun. 2018, 9, 2941. [Google Scholar] [CrossRef]
  108. Hodson, D.J.; Mitchell, R.K.; Marselli, L.; Pullen, T.J.; Gimeno Brias, S.; Semplici, F.; Everett, K.L.; Cooper, D.M.; Bugliani, M.; Marchetti, P.; et al. ADCY5 couples glucose to insulin secretion in human islets. Diabetes 2014, 63, 3009–3021. [Google Scholar] [CrossRef] [Green Version]
Cells 11 03005 g001 550
Figure 1. The applications of metabolomics in diabetic retinopathy. Since 2009, metabolomics studies of DR using various biological samples have become widely available. After sample collection and metabolomics detection and analysis, differential metabolites are obtained, which can be applied to identify biomarkers and explore metabolic targets.
Figure 1. The applications of metabolomics in diabetic retinopathy. Since 2009, metabolomics studies of DR using various biological samples have become widely available. After sample collection and metabolomics detection and analysis, differential metabolites are obtained, which can be applied to identify biomarkers and explore metabolic targets.
Cells 11 03005 g001
Cells 11 03005 g002 550
Figure 2. Strategies for exploring potential molecular targets through metabolomics studies. Twenty-three potential regulatory enzymes (genes) were obtained by integrating metabolomics with GWAS. First, the enzyme-related genes in the disordered metabolic pathways were obtained by retrieving metabolic pathways in the KEGG database. Next, SNPs associated with DM or DR were acquired by searching the GWAS Catalog database. Finally, the enzyme-related genes were matched with genes with SNPs.
Figure 2. Strategies for exploring potential molecular targets through metabolomics studies. Twenty-three potential regulatory enzymes (genes) were obtained by integrating metabolomics with GWAS. First, the enzyme-related genes in the disordered metabolic pathways were obtained by retrieving metabolic pathways in the KEGG database. Next, SNPs associated with DM or DR were acquired by searching the GWAS Catalog database. Finally, the enzyme-related genes were matched with genes with SNPs.
Cells 11 03005 g002
Cells 11 03005 g003 550
Figure 3. The metabolic network of purine metabolism, glycine, serine and threonine metabolism, and sphingolipid metabolism in DR with potential enzyme targets. Schematic overview of the DR-related metabolic pathways including purine metabolism, glycine, serine and threonine metabolism, and sphingolipid metabolism with related enzymes with SNP depicted in different color schemes. Purine metabolism is depicted in blue, glycine, serine and threonine metabolism in red, and sphingolipid metabolism in green.
Figure 3. The metabolic network of purine metabolism, glycine, serine and threonine metabolism, and sphingolipid metabolism in DR with potential enzyme targets. Schematic overview of the DR-related metabolic pathways including purine metabolism, glycine, serine and threonine metabolism, and sphingolipid metabolism with related enzymes with SNP depicted in different color schemes. Purine metabolism is depicted in blue, glycine, serine and threonine metabolism in red, and sphingolipid metabolism in green.
Cells 11 03005 g003
Cells 11 03005 g004 550
Figure 4. The metabolic network of arginine biosynthesis, arginine and proline metabolism, and glutamate metabolism in DR with potential enzyme targets. Schematic overview of the DR-related metabolic pathways and enzyme genes with SNP. Enzymes involved in arginine biosynthesis, arginine and proline metabolism, and glutamate metabolism are depicted in red, green and blue, respectively.
Figure 4. The metabolic network of arginine biosynthesis, arginine and proline metabolism, and glutamate metabolism in DR with potential enzyme targets. Schematic overview of the DR-related metabolic pathways and enzyme genes with SNP. Enzymes involved in arginine biosynthesis, arginine and proline metabolism, and glutamate metabolism are depicted in red, green and blue, respectively.
Cells 11 03005 g004
Table
Table 1. Summary of published studies on metabolomics of diabetic retinopathy.
Table 1. Summary of published studies on metabolomics of diabetic retinopathy.
SpeciesSamplesSubjectsPlatformsDifferential MetabolitesStudy
HumanPlasma64 DR
25 controls		GC–MSArachidonic acid, citric acid, glucose, linoleic acid, l-aspartic acid, methymaleic acid, pyruvic acids, stearic acid, trans-oleic acid, β-hydroxybutyric acidLi et al. (2011) [16]
38 DR
37 controls		HPLC–MS↑: cytosine, cytidine, thynidineXia et al. (2011) [17]
39 DR
35 NDR		UPLC-MS↑: adenosine, inosine, uric acid, xanthineXia et al. (2014) [18]
80 DR
80 controls		GC-MS↑: erythritol, gluconic acid, lactose/cellobiose, mannose, maltose/trehalose, ribose, urea, 1,5-gluconolactone, 2-deoxyribonic acid, 3,4-dihydroxybutyric acid
↓: 1,5-anhydroglucitol		Chen et al. (2016) [19]
52 PDR
72 NPDR
59 NDR		UPLC-MS, GC-MS7 amino acids (asparagine, aspartic acid, glutamic acid, glutamine, glycine, methionine, pyroglutamic acid), 6 organic compounds (citric acid, lactic acid, phosphoric acid, succinic acid, urea, uric acid), 7 carbohydrates (fructose, glucose, myo-inositol, 1,5-anhydroglucitol, 3 saccharides), 11 LysoPCsRhee et al. (2018) [20]
28 NPDR
22 NDR		LC-MSPGF2αPeng et al. (2018) [21]
21 PDR
21 NDR		UPLC-MS63 metabolites (e.g., acetic acid, cytidine sulfite, dihydrouracil, fumaric acid, imidazolone, L-serine, malonic acid, sulfate, uridine, and β-alanine)Zhu et al. (2019) [22]
83 DR
90 NDR		LC-MS126 metabolites (e.g., arginine, acylcarnitine, argininic acid, citrulline, dehydroxycarnitine, glutamic γ-semialdehyde)Sumarriva et al. (2019) [23]
21 PDR
21 NPDR
32 NDR		UPLC-MSAcetylcarnitine, butyryl carnitine, cholic acid, D-glucuronic acid, D-(+)-pantothenic acid, dehydroisoandrosterone sulfate, pantothenic acid, pseudouridine, hypoxanthine, N2,N2-dimethylguanosine, N-acetyltryptophan, leucylleucine, sn-glycero-3-phosphocholine, propionylcarnitine, inosine, urocanic acid, N-fructosyl isoleucine, kynurenic acid, phenylacetylglutamine, glutamine, (−)-riboflavin, 3-methylhistidine,Sun et al. (2021) [24]
64 PDR
92 NPDR
159 NDR		LC-MS↑: arginine, citrullinePeters et al. (2021) [25]
Serum176 DR
329 NDR		LC-MS↑: asymmetric dimethylarginine (ADMA), L-arginine, symmetric dimethylarginine (SDMA)Abhary et al. (2009) [26]
689 DR
216 controls		GC-MS,
LC-MS		12-hydroxyeicosatetraenoic acid (12-HETE) and 2-piperidoneXuan et al. (2020) [27]
43 DR
44 controls		UHPLC–MS↑: 13 lipid (sub)classes (Cers, CerG1s, ChEs, DGs, FAs, LPCs, LPEs, LPC-Os, LPE-ps, PCs, PC-Os, PE-ps, SMs)Xuan et al. (2020) [28]
51 PDR
123 NPDR 143 NDR		LC–MSDR vs. NDR: 62 metabolites
PDR vs. NDR: 53 metabolites
NPDR vs. NDR: 30 metabolites
PDR vs. NPDR: 8 metabolites		Yun et al. (2020) [29]
69 DR
69 NDR		UPLC-MS↑: nicotinuric acid, o-cresol, ornithine, phenylacetylglutamine, p-cresol
↓: alpha-linolenic acid, arachidonic acid, cis-docosahexaenoic acid, gamma-linolenic acid, linolelaidic acid, linoleic acid, palmitoleic acid, cis-7-hexadecenoic acid, hexadecanoic acid, elaidic acid		Zuo et al. (2021) [30]
123 DR
116 NDR		Metabolon DiscoveryHD4Glycoursodeoxycholate, tryptophan, xanthine, phenylacetylglutamine, X-23997, X-13729, 1-palmitoyl-GPA (16:0), and 5-methylthioadenosine (MTA)Yousri et al. (2022) [31]
Erythrocyte70 DR
14 controls		LC-MS↓: arachidonic acid, docosahexaenoic acid, N-6 PUFAs, N-3 PUFAsKoehrer et al. (2014) [32]
Stool45 PDR
90 NDR		UPLC-MSAlantolactone, adenine, corosolic acid, desogestrel, D-erythro-sphinganine, HETE, leukotrieneYe et al. (2021) [33]
21 PDR
14 NDR		UPLC-MS↑: betonicin, butylparaben, traumatic acid, thromboxane B3, salicyluric acid, pyro-L-glutaminyl-L-glutamine, harman, flazine, β-carboline
↓: D-proline, armillaramide, N-gamma-L-glutamyl-D-alanine, N-acetyl-L-methionine, L-threo-3-phenylserine, (R)-pelletierine		Zhou et al. (2021) [34]
Retina20 NPDR
20 NDR		UHPLC-MS↓: long-chain ACs (C ≥ 14), longer-chain FAHFAs, DAGs, TAGs, PCs, CerFort et al. (2021) [35]
Aqueous humor14 DR
13 NDR		NMR↑: asparagine, DMA, glutamine, histidine, threonine
↓: lactate, succinate, 2HB		Jin et al. (2019) [36]
Aqueous and vitreous humor18 PDR
22 controls		LC-MSCysteine persulfides (CysSSH), cystine, oxidized glutathione trisulfide (GSSSG)Kunikata et al. (2017) [37]
Vitreous humor:
28 PDR
22 no diabetes
Aqueous humor:
23 PDR
25 no diabetes		GC-MSVitreous humor:
alanine, alloisoleucine, creatinine, glutamine, leucine, lysine, ornithine, pyroglutamic acid, pyruvic acid, phenylalanine, uric acid, threonine, valine, myoinositol, hydroxylamine;
Aqueous humor:
citrulline, D-glucose, isocitric acid, fructose 6-phosphate, L-lactic acid, threonic acid, myoinositol, D-2,3-dihydroxypropanoic acid		Wang et al. (2019) [38]
Vitreous humor2 PDR
2 PVR
7 no diabetes		NMRunclearYoung et al. (2009) [39]
22 PDR
22 no diabetes		NMR↑: glucose, lactate
↓: ascorbic acid, galactitol		Barba et al. (2010) [40]
16 NPDR
15 PDR
16 no diabetes		LC-MS↑: 5-HETE
↓: 14(15)-EET, 11(12)-EET		Schwartzman et al. (2010) [41]
20 PDR
31 no diabetes		HPLC-MS↑: allantoin, arginine, citrulline, decanoylcarnitine, proline, ornithine, octanoylcarnitine, methionineParis et al. (2015) [42]
9 PDR
8 controls		UHPLC-MSAscorbate, carnitine, citrulline, creatinine, dehydroascorbate, fumarate, glutamine, malate, N-amidino-L-aspartate, sn-glycerol 3-phosphate, proline, pyruvate, tripeptide, ribose, triacanthine, a-ketoglutarate, 5-oxoprolineHaines et al. (2018) [43]
31 PDR
13 no diabetes		LC-MS↑: 5-HETE, 12-HETE, 20-HETE, and 20-COOH-AALin et al. (2020) [44]
35 PDR19 no diabetesUHPLC-MS↑: allantoin, citrulline, dimethylglycine, glycine, lactate, ornithine, pyruvate, proline, urate, N-acetylserine, α-ketoglutarate
↓: creatine, succinate		Tomita et al. (2020) [45]
41 PDR
22 no diabetes		UHPLC-MS↑: 21 oxylipins (ARA, DHA, DTA, EPA, 8S-HETrE, 9-OxoODE, 9S-HOTrE, 9S-HODE, 13S-HOTrE, 13-OxoODE, ±12(13)
-EpOME, 12S-HETE, ±12 (13)-DiHOME, ±9(10)-EpOME, ±9(10)-DiHOME, 13S(γ)-HOTrE, 15-deoxy-Δ12,14-PGJ2, 15S-HETrE, ±14,15-DiHETrE, ±19,20-EpDPE, and 13,14-dihydro PGF2α)		Zhao et al. (2022) [46]
CSF and plasma19 DR
14 controls		NMRAlanine, histidine, leucine, pyruvate, tyrosine, and valineLin et al. (2019) [47]
Plasma and serum228 PDR
276 NPDR 141 NDR		GC-MS, UHPLC-MS↑: 2,4-DHBA, 3,4-DHBA, 3,4-DHBA, ribitol
↓: LPC(16:1), PC(32:1), PC(32:2), TG(50:1), TG(50:2), TG(14:0/16:0/18:1), TG(50:3)		Curovic et al. (2020) [48]
Plasma and
vitreous humor		Plasma:
88 PDR
51 controlsVitreous:
51 PDR
23 controls		UPLC-MS(↑ plasma and vitreous): pantetheine, (24R)-Cholest-5-ene-3-beta,24-diol, alpha-N-phenylacetyl-L-glutamine;
(↓ plasma and vitreous): pipecolic acid;
(plasma ↑, vitreous ↓): pyroglutamic acid		Wang et al. (2022) [49]
Plasma, serum, and urine666 DR
2211 NDR		NMRSerum/plasma: cholesterol esters, creatinine, tyrosine
Urine: citrate, ethanolamine, formate, hypoxanthine		Quek et al. (2021) [50]
RatUrine6 DR rats
6 controls		UPLC-MS↑: cholic acid, kynurenic acid, phenylacetylglycine, p-cresol sulfate, 3-methyldioxyindole, 5-l-glutamyl-taurine
↓: hippuric acid, indoxyl sulfate, p-cresol glucuronide		Wang et al. (2020) [51]
MicePlasma and retina10 db/db mice
10 db/+ mice		LC-MS133 lipids in plasma
61 lipids in retina
15 lipids in plasma and retina (e.g., DAG 34:2, DAG 38:5, LPC 18:1, PC 36:4, SM 36:2)		Sas et al. (2018) [52]
Blood20 db/db mice
10 db/m mice		UHPLC-MSArachidonic acid, cortisol, docosahexaenoic acid, lysoPC (18:0), leukotriene B4, prostaglandin D2, γ-linolenic acidGe et al. (2021) [53]
Zebra-fishwhole body50 pdx1−/− zebrafishUHPLC–MS↑: glutamate, proline, taurine
↓: ornithine, spermidine, tyrosine		Wiggenhauser et al. (2021) [54]
DR, diabetic retinopathy; NDR, no diabetic retinopathy (with diabetes without diabetic retinopathy); PDR, proliferative diabetic retinopathy; NPDR, non-proliferative diabetic retinopathy; PVR, proliferative vitreoretinopathy; GC-MS, gas chromatography mass spectrometry; LC-MS, liquid chromatography mass spectrometry; HPLC-MS, high-performance liquid chromatography mass spectrometry; UPLC-LC, ultra-performance liquid chromatography mass spectrometry; UHPLC-MS, ultra-high-performance liquid chromatography mass spectrometry; NMR, nuclear magnetic resonance; CSF, cerebrospinal fluid.
Table
Table 2. Prediction of potential biomarker of DR in human.
Table 2. Prediction of potential biomarker of DR in human.
SamplesCohortsBiomarkersAUCSensitivitySpecificityStudy
SerumDR VS. NDRA biomarker panel (12-HETE and 2-piperidone)0.9460.8940.919Xuan et al. (2020) [27]
NPDR VS. NDRA biomarker panel (12-HETE and 2-piperidone)0.9580.9290.901Xuan et al. (2020) [27]
DR VS. NDRA biomarker panel (linoleic acid, nicotinuric acid, ornithine, and phenylacetylglutamine)0.9200.9600.780Zuo et al. (2021) [30]
PlasmaDR VS. NDRCytidine0.8490.7370.919Xia et al. (2011) [17]
DR VS. NDRAdenosine0.9130.9471.000Xia et al. (2014) [18]
DR VS. NDR1,5-Gluconolactone, 2-deoxyribonic acid,
gluconic acid, and urea		0.71, 0.68, 0.72, 0.69, respectivelyunclearunclearChen et al. (2016) [19]
DR VS. NDRRatio of the levels of glutamine to glutamic acid0.742unclearunclearRhee et al. (2018) [20]
DR VS. NDRA biomarker panel (alanine, histidine, leucine, pyruvate, tyrosine, and valine)0.836unclearunclearLin et al. (2019) [47]
PDR VS. NDRFumaric acid, uridine, acetic acid, and cytidine0.96, 0.95, 1.00, 0.95, respectivelyunclearunclearZhu et al. (2019) [22]
DR VS. NDRA risk score (pseudouridine)0.8000.9760.531Sun et al. (2021) [24]
PDR VS. (NPDR and NDR)A risk score (pseudouridine, glutamate, leucylleucine and N-acetyltryptophan)0.8200.7620.774Sun et al. (2021) [24]
Vitreous humorPDR VS.
no diabetes		A biomarker panel (galactitol and ascorbic acid)unclear0.8600.810Barba et al. (2010) [40]
PDR VS.
no diabetes		Xanthine, proline, citrulline, pyruvate1.000, 0.986, 0.972, 0.944, respectivelyunclearunclearHaines et al. (2018) [43]
PDR VS.
no diabetes		DTA, EPA, DHA, ARA, ±9(10)-DiHOME,
±19,20-EpDPE, and ±12(13)-EpOME		0.960, 0.803, 0.871, 0.942, 0.805, 0.819, 0.828, respectivelyunclearunclearZhao et al. (2022) [46]
PDR VS.
no diabetes		A biomarker panel (pyroglutamic acid and
pyruvic acid)		0.9510.9550.857Wang et al. (2019) [38]
Aqueous humorPDR VS.
no diabetes		A biomarker panel (D-2,3-dihydroxypropanoic acid, isocitric acid, fructose 6-phosphate, and
L-lactic acid)		0.9650.8800.957Wang et al. (2019) [38]
Cerebrospinal fluidDR VS. NDRA biomarker panel (alanine, histidine, leucine, pyruvate, tyrosine, and valine)0.858unclearunclearLin et al. (2019) [47]
StoolPDR VS. NDRA classifier (Top 5 are alantolactone, desogestrel, adenine, D-erythro-sphinganine, and corosolic acid.)0.9600.8460.936Ye et al. (2021) [33]
AUC, area under the ROC curve; HETE, hydroxyeicosatetraenoic acid; DTA, docosatetraenoic acid; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; ARA, arachidonic acid; DiHOME, dihydroxy-octadecenoic acid; EpDPE, epoxy-docosapentaenoic acid; EpOME, epoxy-octadecenoic acid.
Table
Table 3. Statistics of metabolic pathways associated with DR patients.
Table 3. Statistics of metabolic pathways associated with DR patients.
SamplesPathwaysReported Times
PlasmaPurine metabolism4
Arginine and proline metabolism3
Pyrimidine metabolism3
Alanine, aspartate and glutamate metabolism2
Cysteine and methionine metabolism2
4-hydroxybenzeneacetic acid1
Arachidonic acid metabolism1
Aspartate and asparagine metabolism1
Caffeine metabolism1
Creatinine metabolism1
D-glutamine metabolism1
Fumaric acid metabolism1
Galactose metabolism1
Glyceryl-glycoside metabolism1
Histidine metabolism1
Leukotrienes metabolism1
Linoleic acid metabolism1
Lysine metabolism1
Myo-inositol metabolism1
Niacin metabolism1
Nitrogen metabolism1
Pantothenate and CoA biosynthesis1
Pentose phosphate metabolism1
Phenylalanine metabolism1
Polyol metabolism1
Riboflavin metabolism1
Sphingolipid metabolism1
Sulfur metabolism1
Urea cycle1
α-linolenic acid metabolism1
VitreousArginine and proline metabolism2
Valine, leucine, and isoleucine biosynthesis2
Alanine, aspartate and glutamate metabolism1
Aminoacyl-tRNA biosynthesis1
Glycine and serine metabolism1
Glycolysis1
Nitrogen metabolism1
Pantothenate and CoA biosynthesis1
Pentose phosphate pathway1
Phenylalanine metabolism1
Purine metabolism1
Taurine and hypotaurine metabolism1
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Jian, Q.; Wu, Y.; Zhang, F. Metabolomics in Diabetic Retinopathy: From Potential Biomarkers to Molecular Basis of Oxidative Stress. Cells 2022, 11, 3005. https://doi.org/10.3390/cells11193005

AMA Style

Jian Q, Wu Y, Zhang F. Metabolomics in Diabetic Retinopathy: From Potential Biomarkers to Molecular Basis of Oxidative Stress. Cells. 2022; 11(19):3005. https://doi.org/10.3390/cells11193005

Chicago/Turabian Style

Jian, Qizhi, Yingjie Wu, and Fang Zhang. 2022. "Metabolomics in Diabetic Retinopathy: From Potential Biomarkers to Molecular Basis of Oxidative Stress" Cells 11, no. 19: 3005. https://doi.org/10.3390/cells11193005

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop