Detection of significantly high vitreous concentrations of fatty acid-binding protein 4 in patients with proliferative diabetic retinopathy

The fatty acid-binding protein4 (FABP4) and vascular endothelial growth factor A (VEGFA) play key roles in the metabolic and cardiovascular diseases, and proliferative diabetic retinopathy (PDR), respectively. To identify FABP4 in vitreous fluid in PDR, vitreous concentrations of FABP4 (V-FABP4) and VEGFA (V-VEGFA) from PDR (n = 20) and non-PDR (n = 20) patients were determined by Enzyme-Linked ImmunoSorbent Assays. The data, which included height and weight, systemic blood pressures, several blood biochemical parameters and blood flow at the optic nerve head (ONH) by laser speckle flowgraphy (LSFG) were collected. The levels of V-FABP4 and V-VEGFA were significantly higher in PDR patients than in non-PDR patients (P < 0.001) with a high positive correlation (r = 0.72, P < 0.001) between them. The findings were not affected by body mass index values and the presence of vitreous hemorrhaging. Among the clinical parameters, V-FABP4 correlated positively with creatinine and negatively with age and aspartate transaminase (AST) levels, while V-VEGFA correlated positively with fasting plasma glucose and hemoglobin A1c (HbA1c) levels but negatively with AST. Multiple regression analyses indicated that V-VEGFA, or V-FABP4, AST and HbA1c were independent predictors of V-FABP4 or V-VEGFA, respectively. Both were negatively correlated, but more evident in V-FABP4, with the ONH ocular blood flow.

FABP4, known as adipocyte FABP (A-FABP) or aP2, is expressed in both adipocytes and macrophages, and can be detected in most bodily fluids. Elevated serum concentrations of FABP4 are associated with obesity 8 , insulin resistance 9 , hypertension (HT) 10 , dyslipidemia 11 , atherosclerosis 12 , renal dysfunction 13 , purine metabolism 14 , heart failure and cardiovascular events 15 . Recent studies have also demonstrated that the concentration of FABP4 can be modulated by administering therapeutic drugs for HT, dyslipidemia and DM 9 . These collective observations suggest that FABP4 could also be rationally involved in ocular pathophysiology, especially in DM induced retinopathy. Although another FABP member, FABB5 has been detected within lens so far 16 , our knowledge of the extent of involvement of FABP4 within DR is currently very limited.
In the current study, to elucidate the pathological involvement of FABP4 within the PDR, we surgically collected vitreous specimens from patients with PDR or non-PDR (epiretinal membranes or macular holes) and measured the FABP4 and VEGF concentrations in these samples.

Methods
This study conformed to the principles outlined in the Declaration of Helsinki and was performed with the approval of the institutional ethical committee of our institution. Written informed consent was received from all of the participating subjects.
Patients. Twenty patients who had been consecutively operated on (n = 20 eyes) with PDR (mean age 63.27 ± 11.90 years; 10 male and 10 female, vitreous hemorrhage; 12 eyes, traction retinal detachment; 7 eyes, neovascular glaucoma; 5 eyes, maculopathy; 3 eyes) and 20 patients (mean age 69.2 ± 9.1 years; 8 male and 12 female) with a macular hole (n = 7 eyes) or an epiretinal membrane (n = 13 eyes) requiring vitrectomy were recruited from the Muroran municipal hospital during Jan to Dec, 2017. The 20 former and 20 latter patients were categorized as PDR group and non-PDR group, respectively. In order to determine a suitable surgical indication of vitrectomy, all patients underwent a complete ophthalmologic evaluation before surgery with a bestcorrected visual acuity (BCVA) determination, slit-lamp examination, fundus examination, intraocular pressure measurement, gonioscopy, and optical coherence tomography. A clinical preoperative and intraoperative assessment of disease activity was performed by one experienced retina specialist (K.I). In all patients, under systemic anesthesia, 25 or 27-gauge three-port pars plana vitrectomies were performed (Alcon Constellation Vision System), and simultaneous cataract surgery was added except for 2 eyes each from 20 PDR and 20 non-PDR patient groups. Inter limiting membrane pealing, or air or 10-20% SF6 gas tamponade was performed for 10 PDR eyes and 14 for non-PRD eyes, or 17 PDR eyes and 19 non-PDR eyes, respectively during the surgery. In 12 out of 20 eyes from PDR patients, vitreous hemorrhaging was confirmed prior to the surgery. Post-operatively, no serious complications except for slight vitreous hemorrhaging were observed and none of the eyes have required reoperations as of this writing. Data regarding each patient's general conditions and diabetes control were obtained from the patient and from the patient's general practitioner or diabetologist.
Body height and weight measurements, blood pressure measurements and the collection of peripheral blood specimens for a complete blood count and biochemical analyses were performed as described previously 17 .
Laser speckle flowgraphy (LSFG). The images of the speckle contrast pattern produced by interference as the laser beam was scattered by erythrocytes moving through the ocular fundus vessels were obtained by a fundus camera equipped with an 830 nm diode laser and a charge-coupled device sensor (750 × 360 pixels) (LSFG-NAVI; Softcare Co, Ltd., Fukuoka, Japan) as described previously 18,19 . The LSFG images that were acquired were continuously monitored at 30 frames/sec over a 4-s period and averaged to produce a composite map of ocular blood flow. As a demonstrable indicator of the ocular blood flow at a specific site, the mean blur rate (MBR), as arbitrary units (AU), were calculated and the values at several sites were reconstituted to form a 2-dimensional color-coded map of blood flow velocity. In the current study, we investigated four MBR categories; (1) Average; overall of the optic nerve head (ONH), (2) the vascular area of the ONH (MV) including effects of choroidal vessels, (3) the tissue area of the ONH (MT), and (4) MV-MT (to exclude the effects of choroidal vessels from the MV). All measurements were performed in triplicate and the mean MBR value was calculated. Eye positions were continuously monitored during the LSFG analysis with an auto tracking function, to confirm that the same area was captured again during subsequent examinations. Statistical analysis. Means ± SD for normal distributions or medians (interquartile ranges) for skewed variables were used for expressions of numeric variables. Statistical analyses including (1) intergroup differences by the chi-square test, (2) comparison between two groups by the Mann-Whitney's U test, (3) the distribution of each parameter for its normality using the Shapiro-Wilk W test, (4) logarithmically transformation of non- Table 1. Characteristics of the patients (n = 40). Variables are expressed as number, means ± SD or medians (interquartile ranges). AST aspartate transaminase; ALT alanine transaminase; eGFR estimated glomerular filtration rate; γGTP γ-glutamyl transpeptidase; hsCRP high-sensitivity C-reactive protein; MA mean blur rate of all optic verve head area; MT mean blur rate of tissue area of the optic nerve head; MV mean blur rate of vascular area of the optic nerve head; PDR proliferative diabetic retinopathy. ] pg/mg protein, P = 0.40) were also comparable between genders. Furthermore, it is also known that serum FABP4 levels are elevated by insulin resistance and atherosclerosis, which causes DM and hypertension (HT), respectively. As shown in Table 2, V-FABP4 concentrations were not significantly altered by association with DM among patients with non-PDR and with HT among patients with PDR, although significantly higher V-FABP4 concentrations were observed in the non-PDR patients with HT. In contrast, V-VEGFA concentrations were not significantly altered by association with DM in patients with non-PDR and with HT in patients with PDR or non-PDR (Table 2). These collective results suggest that the origin of V-FABP4 in PDR may be within eyes rather than the peripheral blood circulation, similar to V-VEGFA.   www.nature.com/scientificreports/ observed, but such correlations were more evident in the case of V-FABP4 (Table 4 and Figs. 3 and 4). These data strongly suggest that fluctuations in V-FABP4 levels may be exclusively independent of V-VEGFA, but both could significantly affect the ocular blood circulation.
Stepwise multivariable regression analyses. To further study this, stepwise multivariable regression analyses were assessed using the correlated parameters as possible determinants, as shown in Tables 3 and 4. The data shown in Table 5 demonstrate that Log V-VEGFA, sex and MV-MT, or Log V-FABP4, Log AST and HbA1c were independent predictors of Log V-FABP4 or Log V-VEGFA, respectively, suggesting that V-FABP4 and V-VEGFA independently affect one another.

Discussion
The mechanism responsible for the pathogenesis of DR is known to be extremely complex and involves numerous biochemical and inflammatory processes that are initiated upon long-term exposure to hyperglycemia 21 . During the development of DR, vascular endothelial dysfunction, pericyte loss, and neurodegeneration are simultaneously involved, and ultimately leads to the development of hypoxia and neovascularization 21 . During these progression processes, the local accumulation of cytokines, such as VEGF, tumor necrosis factor-alpha (TNF-⍺), and inducible nitric oxide synthase (iNOS) are substantially induced upon hypoxia in the diabetic retina 22 . This leads also to the accumulation of chemokines and adhesion molecules such as the intercellular adhesion molecule-1 (ICAM-1), and this, in turn, causes the migration of leukocytes towards the retinal endothelium, increased vascular permeability, and the breakdown of the blood-retinal barrier (BRB) 23 . This VEGF related signaling is pivotally involved in the pathogenesis of DR and therefore anti-VEGF therapy becomes an important therapeutic strategy 24 , in addition to classical treatments such as the administration of anti-angiogenic  www.nature.com/scientificreports/ agents 25 or corticosteroids intravitreal injections and laser therapy 26 . However, because of the complexity of the metabolic pathways that are activated during DR, as alluded to above, these single or combined therapies have only limited success [24][25][26] , and therefore additional therapeutic targets independent to the VEGF signaling need to be identified. In the current study, we found that the concentrations of both V-FABP4 and V-VEGFA were substantially elevated in eyes with PDR, and a significantly higher positive correlation (r = 0.72, P < 0.001) was observed between them. However, correlation analyses and stepwise multiple regression analyses for V-FABP and V-VEGFA strongly suggested that both factors were independently regulated, suggesting that V-FABP4 might also be involved in the pathogenesis of PDR. The FABP4, primarily regarded as an adipocyte-and macrophage-specific protein, plays an important role in maintaining glucose and lipid homeostasis 8,15 . The issue of why such high concentrations of adipocyte-and macrophage-specific FABP4 were found in vitreous specimens derived from patients with PDR remains unknown. However, recent studies suggest that FABP4 may be more widely expressed than previously thought, and in fact, FABP4 is also expressed in capillary and venous, but not arterial, endothelial cells under normal conditions 6,7 . The presence of V-FABP4 is not surprising, and suggest that V-FABP4 is most likely derived from retinal capillaries and venous tissue that is affected by PDR. However, the actual origin of the V-FABP4 remains speculative. In addition, a previous study indicated that FABP4 is known to be an indicative biomarker of a general inflammatory degenerative or disease state associated with several metabolic and cardiovascular diseases [8][9][10][11][12][13][14][15] . Therefore, to elucidate origin of the V-FABP4 and their possible pathophysiological roles, further study using suitable animal models will be required to detect what kinds of inflammatory, immune and other cellular signaling are induced.
In terms of relationships between FABP4 and VEGFA, it was reported that VEGFA via VEGF receptor 2 or basic fibroblast growth factor (bFGF) induces the expression of FABP4 in endothelial cells, and in turn, FABP4 www.nature.com/scientificreports/ in endothelial cells promotes angiogenesis 27 . Such an effect of VEGFA on FABP4 expression was inhibited by chemical inhibition or the short-hairpin (sh) RNA-mediated knockdown of the VEGF-receptor-2 (R2), whereas the VEGFR1 agonists, placental growth factors 1 and 2 had no effect on the expression of FABP4 20 . In addition, the knockdown of FABP4 in endothelial cells significantly reduced their proliferation both under baseline conditions and in response to VEGF and bFGF 20 . Alternatively, unlike VEGF, it was reported that the expression of FABP4 in microvascular endothelial cells is induced by cellular senescence and oxidative stress 28 , and is ectopically induced in injured arterial endothelial cells 29 . Since, as above, FABP4 is expressed not only in adipocytes and macrophages but also in several other types of tissues and cells under physiological and pathophysiological conditions, FABP4 may contribute to several aspects of metabolic and cardiovascular diseases as well as renal, respiratory, neurological, gynecological and oncological diseases 8,15 . On the other hand, several drugs, including a statin 30 , eicosatetraenoic acid (EPA)/docosahexaenoic acid (DHA) agent 31 , dipeptidyl peptidase 4 inhibitor (DPP4i) 32 and angiotensin II receptor blocker (ARB) 33 could decrease FABP4 levels.
It is known that angiotensin II (AT II) and components of the renin-angiotensin system (RAS) are expressed in the retina 34 . AT II promotes retinal leukostasis by activating the angiotensin type 1 receptor (AT1-R) pathway that stimulates proinflammatory, proliferative mediators, thus leading to the development and progression of   36 . Selectively blocking the AT1-R, angiotensin receptor blockers (ARBs) has been shown to have neuroprotective and anti-inflammatory effects in animal models with retinal angiogenesis and neovascularization [37][38][39] . In fact, several clinical trials have revealed that inhibiting the RAS by ARB successfully suppressed the incidence and progression of DR 40 . Given the observation that ARB could decrease FABP4 expression as describes above, we rationally speculate that a mechanism involving FABP4 may also contribute to such ARB induced beneficial effects toward DR. To our knowledge, this is the first study to document the presence of V-FABP4 in patients with PDR. However, the current study has several limitations that need to be considered; First, the numbers of patients enrolled in the study were relatively small (n = 40). Nevertheless, in spite of such small numbers in the study groups, we observed a quite strong correlation between V-FABP4 and V-VEGFA (r = 0.72, P < 0.001). Furthermore, elevation of V-VEGFA levels is the consensus observation based on a number of previous studies 41 . Second, current several statistical analyses strongly suggested that V-FABP4 may be involved in the pathogenesis of PDR. However, the precise mechanisms responsible for the pathological contribution of V-FABP4 remains to be elucidated. Therefore, further investigations of the source of the relationship between V-FABP4, V-VEGFA and other related factors within the pathogenesis using larger numbers of patients with PDR with different stages of the disease will also be needed, in addition to in vitro and in vivo studies using animal models as above.