STAT3 activation in circulating myeloid-derived cells contributes to retinal microvascular dysfunction in diabetes
Leukostasis is a key patho-physiological event responsible for capillary occlusion in diabetic retinopathy. Circulating monocytes are the main cell type entrapped in retinal vessels in diabetes. In this study, we investigated the role of the signal transducer and activator of transcription 3 (STAT3) pathway in diabetes-induced immune cell activation and its contribution to retinal microvascular degeneration.
Forty-one patients with type 1 diabetes (T1D) [mild non-proliferative diabetic retinopathy (mNPDR) (n = 13), active proliferative DR (aPDR) (n = 14), inactive PDR (iPDR) (n = 14)] and 13 age- and gender-matched healthy controls were recruited to the study. C57BL/6 J WT mice, SOCS3fl/fl and LysMCre/+SOCS3fl/fl mice were rendered diabetic by Streptozotocin injection. The expression of the phosphorylated human and mouse STAT3 (pSTAT3), mouse LFA-1, CD62L, CD11b and MHC-II in circulating immune cells was evaluated by flow cytometry. The expression of suppressor of cytokine signalling 3 (SOCS3) was examined by real-time RT-PCR. Mouse plasma levels of cytokines were measured by Cytometric Beads Array assay. Retinal leukostasis was examined following FITC-Concanavalin A perfusion and acellular capillary was examined following Isolectin B4 and Collagen IV staining.
Compared to healthy controls, the expression of pSTAT3 in circulating leukocytes was statistically significantly higher in mNPDR but not aPDR and was negatively correlated with diabetes duration. The expression of pSTAT3 and its inhibitor SOCS3 was also significantly increased in leukocytes from diabetic mice. Diabetic mice had higher plasma levels of IL6 and CCL2 compared with control mice. LysMCre/+SOCS3fl/fl mice and SOCS3fl/fl mice developed comparative levels of diabetes, but leukocyte activation, retinal leukostasis and number of acellular capillaries were statistically significantly increased in LysMCre/+SOCS3fl/fl diabetic mice.
STAT3 activation in circulating immune cells appears to contribute to retinal microvascular degeneration and may be involved in DR initiation in T1D.
KeywordsDiabetic retinopathy SOCS3 Monocytes Inflammation
Active proliferative DR
Inactive proliferative DR
Mild non-proliferative diabetic retinopathy
The suppressor of cytokine signalling 3
Signal transducer and activator of transcription 3
Type 1 diabetes
Diabetic microvascular damage affects the eye (retinopathy), kidney (nephropathy) and the nervous system (neuropathy). Although hyperglycemia is the ultimate cause of various diabetic complications, increasing evidence suggests the involvement of inflammation in the initiation and progression of microvascular complications, including diabetic retinopathy (DR) [1, 2]. Inflammation is an adaptive response of the host to endogenous and exogenous insults [3, 4, 5]. During diabetes, hyperglycaemia is a chronic insult to circulating immune cells and vascular endothelial cells. Furthermore, various intermediate metabolic products such as advanced glycation end-products (AGEs) and advanced lipoxidation end-products (ALE) also constitute insults to tissue cells [6, 7, 8]. Vascular endothelial damage is a main pathology in DR and microaneurysms, a characteristic feature of DR, is believed to be the consequence of endothelial and pericyte cell death .
Previous studies have shown that increased leukocyte-endothelial interaction (leukostasis) is a critical step in diabetes-induced retinal microvascular degeneration [9, 10, 11]. Leukostasis is considered as a low-grade intravascular inflammatory response  that occurs from the early stages of diabetes, well before any detectable clinical manifestations have developed [9, 10]. Leukostasis may cause capillary closure, endothelial death and, ultimately, breakdown of the blood-retina barrier (BRB). Activation of both vascular endothelial cells and circulating immune cells is essential for the increased leukostasis in diabetes . Early work in rodent models of DR indicated that neutrophils and monocytes are key immune cells involved in retinal leukostasis [11, 12, 13], and adhesion molecules such as leukocyte function-associated antigen-1 (LFA-1) and vascular cell adhesion molecule 1 (VCAM-1) play an important role in this process . Serra et al. showed that CD11b+ bone marrow-derived monocytes from diabetic mice expressed higher levels of CCR5 , a chemokine receptor for which its polymorphisms have been associated with diabetes . Clinical studies in patients with type 1 diabetes (T1D) have supported this finding, in which monocytes from T1D patients are activated , express higher levels of adhesion molecules such as integrins CD11a and CD11b [17, 18] and have greater potency to bind to endothelial cells .
This study was undertaken to uncover the molecular mechanism(s) underlying diabetes-induced leukocyte activation. The results suggest that the expression of phosphorylated signal transducer and activator of transcription (pSTAT3) is increased in leukocytes from T1D patients with early stages of DR and in the Streptozotocin-induced rodent model of T1D. Deletion of the negative regulator of STAT3, the suppressor of cytokine signalling 3 (SOCS3) in myeloid cells, resulted in increased pSTAT3 expression and uncontrolled diabetes-induced leukocyte activation, increased leukostasis and exaggerated retinal vascular degeneration.
This study was approved by the Office for Research Ethics Committees Northern Ireland (ORECNI, Ref 14/NI/0084); all study procedures were performed in accordance with the Declaration of Helsinki on research involving human volunteers. Participants were recruited from the eye clinics in the Belfast Health and Social Care Trust, Northern Ireland, UK. Written informed consent was obtained from all participants prior to the undertaking of any study procedure.
The study was designed as a cross-sectional study of T1D patients with DR. Participants were included in the study if they were 18 years of age or older, had been diagnosed of having T1D and if they had either mild NPDR or PDR (active or previously treated and inactive). Exclusion criteria included lack of DR; NPDR more than mild or less than PDR; history of severe cardiac diseases, malignancy within the past 5 years, infectious/non-infectious inflammatory diseases within the previous 2 months; presence of active autoimmune disease; history or concurrent use of immunosuppressive medications or steroids; pregnancy; kidney failure and/or media opacities that would prevent adequate fundus examination. All participants underwent a full ophthalmic examination, including detailed slit-lamp biomicroscopy and confirmation and grading of the presence of mild NPDR or PDR. Twenty millilitre whole blood was drawn into EDTA-treated blood collection tube for evaluating STAT3 activation.
Experimental animal studies
Mice and induction of diabetes
C57BL/6 J WT mice, SOCS3fl/fl and LysMCre/+SOCS3fl/fl mice (all in C57BL/6J background) were bred and maintained at the Biological Resource Unit of Queen’s University Belfast. All animals had free access to food and water and were housed in a temperature- and light-controlled environment with 12 h light/dark cycle. All procedures were approved by the Ethics Committee of Queen’s University Belfast, and complied with the Home Office Animal (Scientific Procedures) Act (UK) and the ARVO (Association for Research in Vision and Ophthalmology) statement for the use of animals in ophthalmic and vision research. Diabetes was induced in 12-week male mice by administrating five daily i.p. injections of Streptozotocin (STZ, Sigma-Aldrich, Dorset UK) (50 mg/kg body weight) . Non-fasting glucose levels were tested 1 week after the animals were rendered diabetic using glucometry (FreeStyle Lite, Abbott laboratories, Dublin, Ireland); animals with glucose concentrations > 13 mmol/l were considered to have developed diabetes. Body weight and non-fasting glucose levels were monitored biweekly. HbA1c levels were determined using Glyco-Tek Affinity Column (Helena Biosciences Europe, Gateshead, UK) when mice were sacrificed.
Whole blood RNA extraction and qPCR
Up to 50 μl of mouse whole blood was collected from tail vein and added to RNAprotect blood tube (Qiagen, Manchester, UK) immediately. RNA extraction from whole blood was performed using RNeasy Protect Animal blood kit (Qiagen) following the manufacturer’s instruction. RNA quantity was measured using NanoDrop 1000 Spectrophotometer (Thermo Scientific, Leicestershire, UK). The same amount of RNA from each sample was reverse-transcribed to cDNA using First Strand transcriptase (Invitrogen, Paisley, UK) with random primers following manufacturer’s instruction. Real-time qPCR was performed using Roche SYBR Green Master Mix (Roche, Basel, Switzerland) on Roche LightCycler 480 (Roche). Primers to detect socs3 (forward: cctttgacaagcggactctc, Reverse: gccagcataaaaacccttca) and 18S (aggggagagcgggtaagaga, reverse: ggacaggactaggcggaaca) were designed using primer3 and synthesised by Integrated DNA Technologies (Leuven, Belgium). Data were analysed using delta delta Ct method after normalisation using housekeeping gene 18S.
Whole mouse eyes were collected and fixed in 2% PFA/PBS for 2 h before transferring to PBS and stored in 4 °C for further processing . For retinal flatmount staining, retinal tissues were permeabilised and blocked in 1% BSA/PBS for 2 h and then incubated in primary antibodies (Biotinylated Griffonia Simplicifolia Lectin I Isolectin B4: 1:50, Vector laboratory, Peterborough, UK; rabbit anti-mouse Collagen IV: 1:50, AbD Serotec®, Kidlington, UK) overnight. After thorough washes, samples were incubated with secondary antibodies (Streptavidin FITC or goat anti-rabbit Alex Fluor 569, all from Invitrogen, UK, 1:200) at room temperature for 2 h. Tissues were flatmounted on glass slides with mounting medium and evaluated under confocal microscope (EZ-C1 confocal system, Nikon UK Limited, Surrey, UK).
Leukostasis was evaluated using a protocol published previously with slight modification . Briefly, mice were intravenously injected with 120 μl FITC labelled Concanavalin A (Con A, Vector Lab, UK). After 20 min, mice were deeply anaesthetised with pentobarbital and perfused with PBS under controlled pressure. Eyes were then collected and fixed in 2% PFA for 2 h. Retinas were dissected and flatmounted for confocal microscopy. Adherent Con A+ cells in retinal blood vessels were counted manually.
Flow cytometric analysis in human and mice blood samples
One hundred microliter of human whole blood and 50 μl of mouse blood incubated with/without 100 ng/ml of IL-6 (20 min for human cells, 40 min for mouse cells) were stained for pSTAT3 (Tyr705-FITC from BD biosciences, oxford, UK for human; Tyr705-APC from eBioscience for mouse) following the manufacturers’ instruction. Mouse IgG2a-FITC isotype control (BD Biosciences) or mouse IgG2b Kappa Isotype control (eBioscience) were used. Stained cells were examined by FACS Canto II (BD Biosciences) and data were analysed using the FlowJo software (Version 7, Tree Star, Inc., Ashland, OR, USA).
Thirty microliter of mouse blood were collected through tail-vein prick to EDTA-treated tubes and incubated with antibodies against leukocyte surface markers including CD3, CD19, GR-1, CD11b, LFA-1 (1:100, BD Biosciences), CD62L, (1:100, BioLegend, San Diego, USA) and MHC II (1:100, eBioscience) for 30 min. Red blood cells were then removed with lysis buffer (BD Biosciences) and samples analysed by FACS Canto II (BD Biosciences), and data were analysed using FlowJo (Version 7, Tree Star).
Plasma were collected from control and diabetic mice (with 3 months diabetes duration). Cytokines IL-6, IL-10, TNFα, IFNγ, IL-1β and IL12/IL23p40 as well as chemokine CCL2 were measured by Cytometric Bead Array (CBA; flex set, BD Biosciences) following manufacturer’s instruction. Samples were analysed by FACS Canto II (BD Biosciences) and data were processed by using FCAP array software (BD Biosciences).
Data are expressed as means ± standard error of the mean (SEM). Graphs and statistical analyses were performed using GraphPad Prism 5 (Graphpad, San Diego, CA, USA). Unpaired t test was used to compare two groups; one-way ANOVA with Tukey’s post-hoc test was used to compare three or more groups. Two-way ANOVA was used to compare studies having four groups (SOCS3fl/fl controls, SOCS3fl/fl diabetes, LysMCre/+SOCS3fl/fl controls, LysMCre/+SOCS3fl/fl diabetes). A p value of < 0.05 was considered statistically significant. Statistical Package for the Social Sciences, Windows version 24 (SPSS; SPSS Inc., Armonk, NY, USA) was used to carry out linear regression and correlation of duration of diabetes and pSTAT3 expression.
Increased expression of pSTAT3 in circulating leukocytes in mNPDR
Diabetes duration is a risk factor for development and progression of DR. A statistically significantly negative correlation between durations of diabetes and pSTAT3 expression [baseline, unstimulated expression (Fig. 1d) and IL-6 stimulation (Fig. 1e)] was found. These results suggest that a higher level of pSTAT3 expression in circulating leukocytes is related to early stages of DR and may contribute to the development of DR.
Activation of the IL6-STAT3-SOCS3 pathway in circulating leukocytes in diabetic mice
To elucidate the role of the STAT3 pathway in diabetes-induced immune cell activation and its contribution to the initiation of DR, the STZ-induced type 1 diabetic mouse model was used in the rest of the studies.
STAT3 is activated through phosphorylation of tyrosine 705 (pSTAT3) leading to intranuclear translocation and upregulation of a variety of inflammation/angiogenesis-related genes. Flow cytometry analysis of whole blood after the stimulation of IL-6 showed that the percentage of pSTAT3+ leukocytes (Fig. 2b, c) and the geometric mean of fluorescence intensity of pSTAT3 (Fig. 2d) were significantly higher in diabetic mice compared with control non-diabetic mice. SOCS3 is a negative regulator of pSTAT3. Real-time RT-PCR analysis showed that the expression of SOCS3 was increased in circulating leukocytes from diabetic mice compared with non-diabetic controls (Fig. 2e, p < 0.05). Our results suggest the activation of the IL6-STAT3-SOCS3 pathway in circulating leukocytes in diabetic mice.
Activation of circulating immune cells in diabetic mice
Diabetes-induced myeloid cell activation is heightened in LysMCre/+SOCS3fl/fl mice
Diabetes-induced leukostasis is known to be caused predominately by myeloid cells, e.g. neutrophils and monocytes [10, 11]. To further understand the role of pSTAT3 in diabetes-induced myeloid cell activation, we used LysMCre/+SOCS3fl/fl mice, in which SOCS3 was deleted in myeloid cells resulting in spontaneous STAT3 activation [21, 22, 23]. Under non-diabetic conditions, there was no significant difference between SOCS3fl/fl and LysMCre/+SOCS3fl/fl mice in the distribution of different subsets of circulating leukocytes, including CD3+ T cells, CD19+ B cells, GR-1+ neutrophils, CD11b+ myeloid cells and CD56+ NK cells .
Deletion of SOCS3 in myeloid cells worsens diabetic retinal vasculopathy
The evaluation of acellular capillaries was undertaken using dual staining of Isolectin B4 (for endothelial cells) and collagen IV (for blood vessel basal membrane). Acellular capillaries have basal membrane but no endothelial cells, therefore, are collagen IV positive and Isolectin B4 negative (Fig. 5d–f). A significantly higher number of acellular capillaries were detected in LysMCre/+SOCS3fl/fl diabetic mice than that in the SOCS3fl/fl counterparts (Fig. 5, p < 0.05).
Capillary occlusion is an early event of DR and a hallmark of retinal ischemia in patients with diabetes . Entrapment of leukocytes inside retinal blood vessels (leukostasis) is considered as an important contributory factor to vascular damage, at least in experimental models of DR [24, 25, 26]. Myeloid-derived cells are the predominant cells in the process of leukostasis . Herein, we showed that pSTAT3 expression in circulating immune cells, including myeloid cells, is upregulated in T1D patients with mNPDR but not aPDR, suggesting that STAT3 activation occurs early in DR and may contribute to the development of this complication. STZ-induced diabetic mice is a model representing early stages of DR (e.g. mild vascular degeneration, BRB leakage but no severe ischemia and angiogenesis). Using this model, we found that the IL-6/STAT3/SOCS3 pathway is activated in circulating immune cells. Furthermore, using the LysMCre/+SOCS3fl/fl mice, where the myeloid cells constitutively express pSTAT3 due to the deletion of its inhibitor SOCS3, we showed that uncontrolled STAT3 activation in myeloid cells leads to more severe diabetes-induced retinal vascular changes. Our results suggest that STAT3-driven immune cell activation plays an important role in early diabetes-induced microvascular degeneration.
Signs of mNPDR include microaneurysms and small haemorrhages or transient BRB leakage [27, 28], which are all related to microvascular damage. Previous studies have demonstrated that STAT3 activation in endothelial cells contributes to the early vascular damage in DR [29, 30, 31]. STAT3 may contribute to the early vascular damage through activating genes related to inflammation and angiogenesis such as VEGF and hypoxia-inducible factor 1 [32, 33]. STAT3 activation is also known to be involved in hyperglycaemia-induced endoplasmic reticulum stress in endothelial cells ; inhibiting NOX4 can attenuate high-glucose induced reactive oxygen species (ROS) generation and STAT3 activation . Our study suggests that STAT3 activation in leukocytes also contributes to diabetes-induced vascular damage.
How leukocytes are activated in diabetes is not fully understood. Circulating leukocytes originate from bone marrow (BM) haematopoietic stem cells. A previous study has shown that diabetic microenvironment impairs the colony-forming capability of the BM resulting in increased Ly6Chi (pro-inflammatory) circulating monocytes . In this study, the percentage of CD11b+ or Ly6G+ leukocytes was significantly increased, whereas the percentage of CD3 T cells did not change in the STZ-induced diabetic mice. Further studies will be needed to understand how other leukocytes such as NK cells and B cells are affected by hyperglycemia in this STZ-model. Recently, we have shown that under high glucose conditions, bone marrow-derived macrophages presented a more pro-inflammatory status . It is possible that the high glucose may induce the pro-inflammatory phenotype of macrophages through the STAT3 pathway as STAT3 can be activated by both growth factor receptor and non-receptor tyrosine kinases pathway . Apart from hyperglycaemia, the phenotype and function of circulating cells may also be affected by cytokines, chemokines and other growth factors in the plasma in diabetes. Increased levels of IL-6 and CCL2 have been observed in the plasma  as well as in ocular fluids [38, 39] of patients with DR. We have detected increased levels of IL-6 and CCL2 in the plasma from diabetic mice. Neutrophils and monocytes from diabetic patients produced significantly greater amounts of superoxide [16, 40], and are the main sources of pro-inflammatory cytokines . These cytokines and superoxide may affect circulating immune cells and induce pSTAT3 expression .
STAT3 is a critical transcription activator in physiological and pathological angiogenesis . Previously, we have shown that STAT3 activation in circulating immune cells contributes to the pathogenesis of neovascular age-related macular degeneration . Interestingly, the expression of pSTAT3 in circulating immune cells was not affected in aPDR patients and more importantly, the expression level was negatively correlated with the duration of diabetes. Macrophages are believed to play an important role in pathogenic blood vessels growth in PDR . Our result suggests that the angiogenic effect of macrophage in PDR is not driven by uncontrolled STAT3 activation in their precursors, i.e. monocytes in circulation. Indeed, late-stage diabetes patients often suffer from impaired wound-healing and reduced angiogenesis. However, our result does not rule out the role of STAT3 in PDR as STAT3 can be activated locally in the ischemic retina. When macrophages or other immune cells migrate to the ischemic retina, they may undergo phenotypic switch and express pSTAT3. Cytokines IL-6, IL-17 and growth factor VEGF and PDGF are potent activators of STAT3, and the vitreous levels of these cytokines/growth factors were significantly higher in PDR patients compared with those in cataract patients [44, 45, 46]. The role of pSTAT3 in the pathogenesis of PDR warrants further investigation.
This study uncovered the novel role of the IL-6/STAT3 pathway in diabetes-induced circulating immune cell activation, and the development of diabetic retinal vasculopathy. Targeting the IL-6/STAT3 pathway in myeloid cells may be a novel approach for the prevention or the treatment of retinal capillary occlusion in DR patients.
The authors would like to thank Miss Elizabeth Sloan for her generous support.
MC designed and executed animal experiments, collected data and drafted the manuscript. GO executed human study, collect data. DA and NJL recruited patients. AK discussed and interpreted data. NL recruited patients and interpreted data. HX designed experiments, interpreted data and drafted the manuscript. All authors read and approved the final manuscript.
The study is funded by Juvenile Diabetes Research Foundation (JDRF, Ref 2-SRA-2014-141-Q-R), Fight for Sight (1574/1575).
Ethics approval and consent to participate
Human study was approved by the Office for Research Ethics Committees Northern Ireland. Written consent was obtained from all participants. The use of animals was approved by the Home Office UK and the Animal Welfare and Ethical Review Body at Queen’s University Belfast.
Consent for publication
Consent was obtained from all participants in human study.
The authors declare that they have no competing interests.
- 11.Serra AM, Waddell J, Manivannan A, Xu H, Cotter M, Forrester JV. CD11b+ bone marrow-derived monocytes are the major leukocyte subset responsible for retinal capillary leukostasis in experimental diabetes in mouse and express high levels of CCR5 in the circulation. Am J Pathol. 2012;181(2):719–27.PubMedCrossRefGoogle Scholar
- 21.Qin H, Yeh WI, De Sarno P, Holdbrooks AT, Liu Y, Muldowney MT, et al. Signal transducer and activator of transcription-3/suppressor of cytokine signaling-3 (STAT3/SOCS3) axis in myeloid cells regulates neuroinflammation. Proc Natl Acad Sci U S A. 2012;109(13):5004–9.PubMedPubMedCentralCrossRefGoogle Scholar
- 23.Chen M, Zhao J, Ali IHA, Marry S, Augustine J, Bhuckory M, et al. Cytokine signaling protein 3 deficiency in myeloid cells promotes retinal degeneration and angiogenesis through Arginase-1 up-regulation in experimental autoimmune Uveoretinitis. Am J Pathol. 2018;188(4):1007–20.PubMedCrossRefGoogle Scholar
- 29.Li J, Wang JJ, Yu Q, Chen K, Mahadev K, Zhang SX. Inhibition of reactive oxygen species by lovastatin downregulates vascular endothelial growth factor expression and ameliorates blood-retinal barrier breakdown in db/db mice: role of NADPH oxidase 4. Diabetes. 2010;59(6):1528–38.PubMedPubMedCentralCrossRefGoogle Scholar
- 30.Chen Y, Wang JJ, Li J, Hosoya KI, Ratan R, Townes T, et al. Activating transcription factor 4 mediates hyperglycaemia-induced endothelial inflammation and retinal vascular leakage through activation of STAT3 in a mouse model of type 1 diabetes. Diabetologia. 2012;55(9):2533–45.PubMedPubMedCentralCrossRefGoogle Scholar
- 34.Hazra S, Jarajapu YP, Stepps V, Caballero S, Thinschmidt JS, Sautina L, et al. Long-term type 1 diabetes influences haematopoietic stem cells by reducing vascular repair potential and increasing inflammatory monocyte generation in a murine model. Diabetologia. 2013;56(3):644–53.PubMedCrossRefGoogle Scholar
- 37.Koleva-Georgieva DN, Sivkova NP, Terzieva D. Serum inflammatory cytokines IL-1beta, IL-6, TNF-alpha and VEGF have influence on the development of diabetic retinopathy. Folia Med (Plovdiv). 2011;53(2):44–50.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.