An oral absorbent, AST-120, Restores Vascular Growth and Blood Flow in Ischemic Muscles in Diabetic Mice via Modulation of Macrophage Transition

Diabetes has a pronounced effect on the peripheral vasculature. The accumulation of advanced glycation end products (AGEs) is regarded as the crucial mechanism responsible for vascular damage in diabetes, but it is not easy to be avoided from food. In this study, we aimed to investigate the effects of an oral absorbent, AST-120, on the accumulation of AGEs and changes in blood ow recovery in diabetic mice. into wild-type (WT) mice WT with mixed into with pulverized mellitus and to

Cardiovascular complications are the leading cause of morbidity and mortality in patients with diabetes mellitus. Diabetes has a pronounced effect on the peripheral vasculature, which results in high nontraumatic amputation rates. The management of peripheral arterial disease in patients with diabetes is challenging, and the outcomes of revascularization are poor [1]. Patients with diabetes have diminished capacity for neovascularization to overcome the reduction in blood ow following arterial stenosis or occlusion [2]. This probably contributes to the adverse outcomes when limb ischemia develops in these patients. Few studies have speci cally evaluated the effects of diabetes on angiogenesis in peripheral vascular diseases. Furthermore, the mechanisms underlying the impaired neovascularization remain largely unde ned, resulting in limited therapeutic strategies to improve critical limb ischemia.
The accumulation of advanced glycation end products (AGEs) is a crucial mechanism responsible for the vascular damage in diabetes [3][4][5]. AGEs are post-translational modi cations of proteins in which amino acids and carbohydrates interact via an irreversible, non-enzymatic process. Chronic hyperglycemia and the altered redox state in diabetes result in increased formation of AGEs. One mechanism by which the effects of AGEs are elicited is through their binding with the receptor for advanced glycation end products (RAGE) [6][7][8]. Binding of AGEs with RAGE activates several intracellular pathways, resulting in the upregulation of multiple pro-in ammatory genes [9]. In diabetes, RAGE has been demonstrated to be elevated and co-localized with AGEs in a variety of tissues, especially in the affected vasculature [10]. Activation of RAGE has also been observed in endothelial cells, vascular smooth muscle cells, and macrophages [11], which play signi cant roles in neovascularization following tissue ischemia.
AST-120 (Kremezin®) is an oral absorbent that can bind to many low-molecular-weight (100-10,000 kDa) compounds, including carboxymethyl lysine (CML), a well-characterized AGE [12].  has been demonstrated to decrease the serum levels of AGEs in patients with chronic kidney disease (CKD) but without diabetes [12][13][14]. We hypothesized that AGEs and RAGE signaling could play inhibitory roles in angiogenesis by promoting excessive in ammation in ischemic limbs of patients with diabetes.
To test this hypothesis, we investigated the effects of treatment with AST-120 on the accumulation of AGEs in diabetic mice and evaluated its impact on in ammation, neovascularization, and reperfusion in a hind-limb ischemia model.

Methods
Food Preparation and AGE-AST120 Binding Assay Regular pulverized chow food was prepared without heat exposure (Regular-AGE); it contained 18% protein, 58% carbohydrate, 7.5% fat, and 3.73 kcal/g. Cooked pulverized chow food was prepared by heating at 125 o C for 60 min. Based on the measurement by an AGE-sensitive enzyme-linked immunosorbent assay (ELISA), this preparation contained 200 ng/ml of AGEs (Fig. 1B) and was used as a high-AGE diet. The concentration of CML demonstrated a similar pattern of two-fold difference between regular chow food and cooked chow food (Fig. 1C). To identify the binding effect of AST-120 in regular chow food or cooked chow food, both foods were individually mixed with 5% AST-120 (Kremezin, Kureha Corporation, Osaka, Japan) in water. After 16 hours of incubation at 37 o C, the supernatant was measured using AGE and CML ELISA kits.

Animals
Eight-week-old male wild-type (WT) FVB mice were used in this study. The mice were divided into four groups based on the food they received (Fig. 1A): regular pulverized chow food (Chow food), 5% AST-120 mixed with regular chow food (Chow food + AST-120), cooked pulverized chow food (Cooked chow food), and 5% AST-120 mixed with cooked chow food (Cooked chow food + AST-120). After six weeks of the food consumption, the mice were euthanized., and their plasma samples were obtained for AGEs and CML measurements. In this experiment, intraperitoneal injection of tribromoethanol (avertin; 225-240 mg / kg body weight) was used as the anesthetic agent. After con rming the mouse in anesthesia, it was sacri ced by bleeding.
The mice were rendered diabetic by intraperitoneal injection of saline or streptozotocin (STZ: 5 days, 50 mg/kg in citrate buffer, 100 mmol/L, pH 4.5; Sigma, St. Louis, MO, USA). Blood glucose was measured 7 days after the rst STZ injection using a blood glucose meter. Animals with blood glucose levels > 250 mg/dL were considered to be diabetic and were included for further experiments. These mice were divided into four groups: WT mice with no treatment (WT), WT mice treated with 5% AST-120 mixed with pulverized chow (WT + AST-120), mice with STZ-induced diabetes mellitus (DM), and mice with STZinduced diabetes mellitus treated with 5% AST-120 (DM + AST-120) (Fig. 1F). To con rm that the STZinduced DM mice represented a model of human type 2 diabetes, we performed blood glucose tests and intraperitoneal glucose tolerance tests (IPGTTs).

AGE and CML Determination
Circulating levels of AGEs and CML were determined using competitive ELISA kits.

Hind-limb Ischemia Model
Hind-limb ischemia was induced by excising the right femoral artery, as described previously [15]. The animals were anesthetized with intraperitoneal ketamine (100 mg/kg) and xylazine (10 mg/kg). The proximal and distal portions of the right femoral artery and the distal portion of the right saphenous artery were ligated. Subsequently, the arteries and all the branches were dissected free and excised.

AGEs and Cytokine Assays
Before hind-limb ischemia surgery, blood samples were collected from the facial veins of the mice.
Plasma concentrations of AGEs were determined using a commercial enzyme-linked immunosorbent assay (Cell Biolabs, San Diego, CA, USA) according to the manufacturer's instructions. Two weeks after the hind-limb ischemia surgery, the concentrations of plasma cytokines and muscle extracted proteins, including tumor necrosis factor (TNF)-α, interleukin (IL)-6, chemokine (C-X-C motif) ligand 2 (Cxcl2), monocyte chemoattractant protein (MCP)-1, plasma macrophage colony-stimulating factor (M-CSF), vascular endothelial growth factor (VEGF), IL-4, and IL-10 were determined using the commercially available Procarta multiplex immunoassay kit (eBioscience, San Diego, CA, USA) according to the manufacturer's instructions.

Laser Doppler Imaging
The ratio of blood ow in the ischemic (right) limb to that in the non-ischemic (left) limb was measured with a laser Doppler perfusion imaging system (Moor Instruments Limited, Devon, UK). The mice were monitored using serial scanning of surface blood ow in the hind-limb before and after the hind-limb ischemia surgery, which was repeated weekly for four weeks.

RAGE Ligand Distribution, Macrophage In ltration, and Characterization
Immunohistochemistry was performed using goat anti-human RAGE (1:200 dilution; Genetex, CA, USA). Macrophages with the M1 phenotype were detected using rabbit inducible nitric oxide synthase (iNOS) antibody (1:500 dilution), and those with the M2 phenotype were detected using rabbit Arg1 antibody (1:100 dilution) in combination with the mouse macrophage marker F4/80 (1:100 dilution; all products from Abcam). Fluorescently conjugated secondary antibodies were used for signal detection. The M1 and M2 macrophages were counted in ve randomly selected elds of each muscle section. Photographs of the stained histological sections were taken using a confocal microscope (Olympus Micro, Japan) and processed with the FV10-ASW 4.0 imaging system.

Macrophage Cell Cultures
To study the effects of AGEs on macrophage polarization, bone marrow cells were cultured and differentiated into macrophages of different phenotypes ( Supplementary Fig. 1). Bone marrow cells were extracted from 6-week-old WT mice and cultured with M-CSF containing Royal Park Memorial Institute (RPMI) medium every other day. Well-differentiated bone marrow-derived macrophages (BMDMs) were harvested on day 5. BMDMs were incubated for 48 hours with different media: control medium, AGE (100 µg/ml), AST-120-pretreated AGE, high glucose (25 mmol/L D-glucose), high glucose + AGE, and high glucose + AST-120-pretreated AGE. Total RNA was isolated, and quantitative real-time polymerase chain reaction (PCR) was performed. The primers of IL-1b, iNOS, YM-1, Arginase-1 ( Supplementary Fig. 2), and the mixture were purchased from SYBR Green systems (Roche LC480). BMDMs were triggered into M1 and M2 macrophages; IL-1b and iNOS are highly expressed in M1 macrophages, while arginase-1 and YM-1 are highly expressed in M2 macrophages. Data were calculated using the comparative Ct method (△△Ct) and expressed as the -fold increase over the indicated controls.

Statistical Analysis
Data are presented as mean ± standard error of mean. Statistical analysis was performed using the Student's unpaired t-test or one-way analysis of variance (ANOVA), followed by post-hoc Duncan's multiple-comparison test. Analyses were performed using SPSS version 16.0 (SPSS Inc., Chicago, IL, USA). P-values < 0.05 were considered statistically signi cant.

AST-120 Reduces Plasma AGEs via Inhibition of Absorption of Dietary AGEs
The content of AGEs in cooked chow is higher than that in the regular chow. AST-120 decreased the content of AGEs in both regular and cooked chows (Fig. 1B) as well as the content of CML (Fig. 1C). Therefore, AST-120 could adsorb AGEs and CML in the food. Plasma levels of AGEs and CML in mice that were fed cooked chow (high AGE content) were higher than those in mice that were fed regular chow (low AGE content). AST-120 reversed the elevated AGEs and CML levels in mice that were fed both regular chow and cooked chow (Fig. 1D, 1E). These observations con rm the role of dietary AGEs on the circulating AGE and CML levels and the effect of AST-120 therapy.
AST-120 Reduced the Plasma AGE Concentration and Cytokine Levels in Diabetic Mice DM and DM + AST-120 mice had increased blood glucose levels under non-fasting conditions ( Supplementary Fig. 3A) and on IPGTTs ( Supplementary Fig. 3B) when compared with WT mice. Plasma AGE levels were elevated in DM mice when compared with those in WT mice; the AGE level was 67% lower in the DM + AST-120 group than that in the DM group (Fig. 1G). No differences in body weight were observed between the WT and WT + AST-120 groups ( Supplementary Fig. 3C). Therefore, AST-120 reduced the plasma level of AGEs, but not the blood glucose level.
We assessed the expressions of circulating pro-in ammatory, anti-in ammatory, and angiogenic cytokines following the hind-limb ischemia surgery. Two weeks after the surgery, the plasma MCP-1 level was signi cantly higher in DM mice when compared with WT mice. In DM + AST-120 mice, a trend toward lower plasma MCP-1 level was observed when compared with DM mice (Fig. 1H). On the other hand, the levels of IL-10 ( Fig. 1I) and VEGF (Fig. 1J) were higher in WT mice when compared with DM mice. The DM + AST-120 group demonstrated a trend toward higher levels of plasma IL-10 and VEGF when compared with the DM group. There were no signi cant changes in the levels of other plasma cytokines, including TNF-α, IL-6, IL-4, and Cxcl2 (data not shown).

AST-120 Treatment Reversed Blunted Perfusion Recovery and Impaired Neovascularization in the Ischemic Muscles of DM Mice
Laser Doppler perfusion images demonstrated gradual reperfusion of the ischemic limbs with approximately 95% of the blood ow in the non-ischemic limbs at 4 weeks in WT mice ( Fig. 2A). The blood ow recovery was 44% lower by 4 weeks in DM mice when compared with WT mice (P < 0.05) (Fig. 2B). In DM + AST-120 mice, the blood ow ratio in the ischemic limbs recovered signi cantly to the levels observed in the WT mice.
Two weeks following the hind-limb ischemia surgery, the densities of CD31 + capillaries (Fig. 2C, 2D) and α-SMA + small arteries (Fig. 2E, 2F) were signi cantly augmented in the ischemic limbs in WT mice, whereas no such increase was observed in DM mice. DM + AST-120 mice demonstrated signi cant recovery of CD31 + capillaries and α-SMA + small arteries.

Distribution of AGEs and RAGE in the Ischemic Limbs of DM Mice
Our initial data suggested that neovascularization mechanisms were impaired in diabetes and restored, at least in part, by AST-120-mediated reduction of AGEs. To further examine this hypothesis, we quanti ed the number of CML-and RAGE-positive cells in ischemic muscles 2 weeks after the hind-limb ischemia injury (Fig. 3A-3D). Immuno uorescence staining revealed signi cantly higher numbers of CML-positive cells and RAGE-positive cells in DM mice than those in WT mice. Compared with DM mice, the number of CML-positive cells and AGE-positive cells in the ischemic limbs had signi cantly decreased after AST-120 treatment.

Macrophage In ltration in the Ischemic Limbs of DM Mice
Macrophage in ltration between the myocytes was observed in ischemic tissues. DM mice demonstrated a signi cant increase in the total number of F4/80-positive macrophages when compared with WT mice. To clarify the impact of AGEs on the in ltration of macrophages, we assessed the co-localization of F4/80 and RAGE (Fig. 3E) using immuno uorescence staining and confocal microscopy images. The images demonstrated co-localization of F4/80 (yellow), RAGE (red), and the nuclear marker DAPI (4',6diamidino-2-phenylindole; blue) in the same cells in ltrated between the myocytes ( Supplementary  Fig. 4). DM mice demonstrated a signi cantly higher number of RAGE-positive macrophages in the ischemic muscles when compared with WT mice. After 5 weeks of AST-120 treatment, RAGE-positive macrophages had reduced signi cantly in DM + AST-120 mice when compared with DM mice (Fig. 3F).
Effects of Diabetes and AST-120 on M1 and M2 Macrophage Polarization We quanti ed the expression of M1 in comparison with M2 markers following the hind-limb ischemia surgery because these markers are generally associated with pro-in ammatory and pro-angiogenic signatures, respectively [16,17]. The ischemic limbs were stained for F4/80 and M1 (iNOS) and M2 (Arg1) markers. The data were analyzed by assessing the percentage of F4/80-positive macrophages (total macrophage number) that were double-stained with respective M1 or M2 markers under each experimental condition (Fig. 4A, 4B). The percentage of iNOS + macrophages was signi cantly higher in DM mice than in WT mice (Fig. 4C), and it was signi cantly lower in DM + AST-120 mice than in DM mice (Fig. 4D). No difference was observed between WT and WT + AST-120 mice.
Analysis of M2 markers revealed a different pattern. The percentage of Arg1 + macrophages was signi cantly lower in DM mice than that in WT mice. No difference was observed between WT and WT + AST-120 mice. However, DM + AST-120 mice demonstrated a trend toward a higher percentage of Arg1 + macrophages when compared with DM mice (Fig. 4D). Collectively, the predominance of proin ammatory M1 macrophages was attenuated, as opposed to pro-angiogenic M2 macrophages, and vascular growth was restored by AST-120 treatment following the hind-limb ischemia in diabetic mice.
Analogous ndings were found when the data were analyzed according to the ratio of iNOS + cells and the total number of cells in the ischemic limb.
Pro-in ammatory, Anti-in ammatory, and Angiogenic Cytokine Expression in the Ischemic Limbs of DM + AST-120 Mice Two weeks after the hind-limb ischemia surgery, the levels of TNF-α, IL-6, Cxcl2, and MCP-1 protein were signi cantly increased in the ischemic tissues of DM mice when compared with those in the WT mice ( Fig. 5). In the AST-120 treated DM mice, the up-regulated levels of TNF-α, IL-6, and MCP-1 protein were reversed, and the level of Cxcl2 was partially reversed. The levels of IL-10 and VEGF were signi cantly decreased in the DM groups; however, AST-120 treatment did not revere these effects.

Effect of AGEs on Macrophage Polarization in BMDMs:
The binding assay demonstrated that AST-120 could adsorb and decrease the content of AGEs in the culture media (Fig. 6A, 6B). Macrophages treated with AGEs had signi cant increase in iNOS mRNA transcription following incubation for 48 hours. In contrast, AST-120 pre-treated AGEs group had a signi cant decrease in iNOS mRNA transcription when compared with the AGE-alone group (Fig. 6C). In case of M2 markers, AGEs signi cantly downregulated Arg1 mRNA transcription. AST-120 pretreated AGEs reversed the decrease in Arg1 mRNA when compared with the AGEs-alone group (Fig. 6D). To mimic a diabetic condition, these experiments were repeated after incubation in high-glucose medium for 48 hours (Fig. 6E). High-glucose incubation resulted in an increase in iNOS mRNA transcription when compared with that in the control medium. The effects of AGEs and pre-treated AGEs in the high-glucose condition were similar to those in the normal-glucose condition: AGE-treated BMDMs had higher iNOS mRNA transcription when compared with macrophages treated with glucose alone. Pre-treatment with AST-120 partially reversed the change in iNOS mRNA transcription induced by AGEs.

Discussion
The main ndings of this study can be summarized as follows: (1) oral AST-120 reduced the plasma levels of AGEs, probably, via absorption of dietary AGEs; (2) AST-120 therapy decreased circulating and tissue levels of AGEs and resulted in improvements in neovascularization and blood ow recovery in a hind-limb ischemia model in diabetic mice; and (3) the polarity of macrophages in the ischemic tissues was predominantly pro-in ammatory M1 phenotype in diabetic mice, and AST-120 therapy reversed this polarity into the pro-angiogenic M2 phenotype.
In patients with diabetes, neovascularization is insu cient to overcome the loss of blood ow that occurs due to arterial narrowing or occlusion [2]. Angiogenesis-de ned as the sprouting of new blood vessels from pre-existing vascular structures-is a physiological reaction to tissue ischemia [18]. This process starts with the degradation of non-brillar collagens in the basement membrane, followed by migration and proliferation of pre-existing vascular endothelial cells or circulating endothelial progenitor cells (EPCs) [19]. Angiogenesis is initiated by hypoxia and in ammation, and it involves angiogenic, antiangiogenic, and maturation factors [20]. Previous studies have found that alterations in VEGF expression, attenuation of monocyte migratory ability, and impaired EPC mobilization contribute to impairments in neovascularization in diabetes [21,22]. Formation of AGEs is one of the major mechanisms responsible for vascular damage in diabetes [23]. In RAGE knockout mice, collateral growth and blood ow recovery are reduced when compared with WT mice [24]. Blockade of the formation of AGEs by aminoguanidine also helps in the restoration of perfusion in ischemic tissues [25]. Our data are, broadly, in line with these ndings; however, they also provide novel insights into the pathological mechanisms and possible therapeutic approaches.
Many agents have been developed to attenuate the damage induced by AGEs, including inhibitors and breakers of AGEs, antioxidants, natural substances, and anti-in ammatory molecules [23]. Diet-derived AGEs are essential sources of AGEs, and the removal of exogenous AGEs is a potential approach to reduce serum levels of AGEs. It has been demonstrated in healthy subjects that dietary restriction alone could reduce serum levels of AGEs by 30-40% [26,27]. In patients with diabetes or renal failure, restriction of dietary AGEs could attenuate tissue injury related to AGEs [28,29]. AST-120 is an oral adsorbent that can bind to many low-molecular-weight (100-10,000 kDa) compounds. It has been used to reduce a variety of uremic toxins in patients with CKD, with concomitant improvements in the intimalmedial thickness and ow-mediated dilatation [30,31]. AST-120 could completely adsorb CML, a wellrecognized food-derived AGE, and reduce serum levels of AGEs in patients with CKD [12]. For the rst time, we demonstrated that AST-120 could also reduce serum levels of AGEs in diabetic mice with normal renal functions. Circulating AGEs were reduced by approximately 67% in diabetic mice treated with AST-120, resulting in a serum level equal to that in non-diabetic mice. Although the adsorptive property of AST-120 was not examined in gastric or intestinal uids, in vitro data suggest that AST-120 could adsorb AGEs in food and, subsequently, decrease the serum levels of AGEs. The evidence collectively suggests that adsorption of exogenous AGEs may be a potential therapeutic approach to attenuate ischemic injury in diabetes. Further studies are required to clarify the optimal dose of AST-120 and the degree of reduction in the levels of AGEs required to produce clinical bene ts.
Few studies have addressed the underlying mechanisms via which AGEs attenuate neovascularization. Tamarat et al. demonstrated that the formation of AGEs reduced degradation of the extracellular matrix degradation and, subsequently, abrogated the angiogenic process in diabetic mice [25]. Tanii et al. demonstrated that AGEs disturb the recruitment and functions of pericytes via the platelet-derived growth factor-BB/protein kinase C axis, which regulates the maturation of capillary vessels [32]. Our ndings demonstrated that the deleterious effects of AGEs on neovascularization might be related to in ammatory activation in ischemic tissues through the binding of AGEs with RAGE. Macrophages are the principal cells that participate in the in ammatory and angiogenic processes following tissue ischemia. We demonstrated a signi cant increase in the macrophage in ltration in ischemic tissues in diabetic mice, which was reversed by reducing the AGEs with AST-120 therapy. The in ltration by macrophages was accompanied by increases in pro-in ammatory cytokines and decreases in proangiogenic cytokines. The effects of AGEs on macrophages may be mediated through the binding to RAGE, as demonstrated by the co-expression of RAGE and F4/80 positive cells in ischemic tissues. AST-120 decreased not only the circulating AGEs but also the AGE-positive cells, RAGE-positive cells, and RAGE as well as F4/80 double-positive cells in ischemic tissues. The detailed mechanism of enhanced activation of macrophages by AGEs, mainly via the RAGE/NF-κB pathway was demonstrated by Jin et al. [33].
In addition to in ammatory activation, macrophages respond to environmental signals and transform into different functional phenotypes, ranging from pro-in ammatory (classic M1 activation) to proangiogenic phenotypes (alternative M2 activation) [34,35]. The pro-angiogenic property is not demonstrated in all subsets of macrophages. M2 macrophages promote angiogenesis by producing proangiogenic cytokines and growth factors. In diabetic mice, we found the pro-in ammatory M1 phenotype to be the predominant one in ischemic tissues. In non-diabetic mice, the predominant macrophages were of the pro-angiogenic M2 phenotype. A comparable change in pro-in ammatory cytokines (TNF-α, IL-6, MCP-1, and Cxcl6) and pro-angiogenic cytokine (VEGF) expression supports the role of macrophage polarization in the angiogenic process. AST-120 reversed the M1/M2 polarization, resulting in subsequent improvements in neovascularization. The effects of AGEs on macrophage polarity were demonstrated in in vitro studies as well. Either AGEs or high-glucose conditions enhance the transformation of macrophages into M1 phenotype. Administration of AST-120 could reverse the change in polarity.
The impact of macrophage polarization on the angiogenic process was investigated in previous studies.
Angiogenic growth factors and cytokines are highly expressed in M2 rather than M1 macrophages, and M2 macrophages promote tube formation by these secretory factors in vitro. [36,37]. Neutralizing these growth factors could impair the M2-induced angiogenesis [37]. In animal models, the infusion of antiin ammatory M2 macrophages but not pro-in ammatory M1 macrophages promotes angiogenesis [37].
Collectively, AGE-RAGE related alternations in macrophage polarization might be a probable mechanism underlying the microangiopathy in diabetes.
Some limitations of this study must be acknowledged. First, AGEs are recognized as some of the most potent toxins in diabetes that can be reduced by AST-120. Although mice without CKD were used in this experiment, we could not exclude the possibility that the bene ts were derived from reductions in other toxins. Second, the causal relationship between AGEs and neovascularization and detailed mechanical pathways involved could not be proven directly in this animal model.

Conclusions
AGEs may have deleterious effects on neovascularization by altering the polarization of macrophages from the angiogenic to the pro-in ammatory phenotype in diabetes. Administration of AST-120 to diabetic mice improved neovascularization and blood ow recovery in ischemic tissues, probably, by decreasing the intestinal absorption of AGEs. AST-120 may be an effective novel therapeutic agent to facilitate neovascularization and improve the outcomes in patients with diabetic and peripheral arterial disease.  AST-120 reduce dietary advanced glycation end products (AGEs) absorption in diabetic mice. (A) Mice were divided into four groups: mice treated with regular pulverized chow food (Chow food), mice treated with 5% AST-120 mixed into regular chow food (Chow food + AST-120), mice treated with cooked pulverized chow food (Cooked chow food), and mice treated with 5% AST-120 mixed into cooked chow food (Cooked chow food + AST-120). To invested the binding effect of AST-120, the supernatant from four group's food was measured by AGE (B) and CML (C) ELISA kits. After 6 weeks treatment, mice plasma AGE (D) and CML (E) levels were measured. (F) To study the AST-120 treatment in diabetic mice, mice were divided into four groups: WT, WT + AST-120, DM, and DM + AST-120. Plasma AGE (G), MCP-1 (H), IL-10 (I), and VEGF (J) levels were measured 14 days after hindlimb ischemia surgery. *P < 0.05 vs.

Abbreviations
WT, **P < 0.01 vs. WT, #P < 0.05 vs. DM; n = 11 per group.    Patterns of macrophage-related cytokine production in ischemic muscles of mice. On day 14 after injury, tissue extract was made from ischemic mouse muscle, and the following protein levels were detected.