Inhibiting adipose tissue M1 cytokine expression decreases DPP4 activity and insulin resistance in a type 2 diabetes mellitus mouse model

Adipose tissue inflammation is a major cause of the pathogenesis of obesity and comorbidities. To study the involvement of M1/M2 cytokine expression of adipose tissue in the regulatory mechanisms of dipeptidyl peptidase 4 (DPP4) and insulin resistance in diabetes, stromal vascular fractions (SVFs) were purified from inguinal adipose tissue of diabetic (Leprdb/db) and non-diabetic (Lepr+/+) mice followed by analysis of M1/M2 cytokine expression. SVFs of Leprdb/db mice exhibited increased TNF-α, IL-6, IL-1β, CCL2, and DPP4 mRNA expression but decreased IL-10 mRNA expression. Plasma from Leprdb/db mice induced TNF-α, IL-6, IL-1β, CCL2, and DPP4 mRNA expression and plasma from Lepr+/+ mice induced IL-10 mRNA expression in SVFs from Leprdb/db mice. Injection of Lepr+/+ plasma into the adipose tissue of Leprdb/db mice decreased mRNA expression of TNF-α, IL-6, IL-1β, CCL2, and DPP4 and protein expression of pJNK and DPP4 in SVFs, reduced mRNA expression of ICAM, FMO3, IL-1β, iNOS, TNF-α, IL-6, and DPP4 and protein expression of ICAM, FMO3, and DPP4 in liver, and suppressed mRNA expression of TNF-α, IL-6, IL-1β, and DPP4 in Kupffer cells. Plasma from Leprdb/db mice did not induce M1 cytokine expression in SVFs from Leprdb/db-Jnk1-/- mice. Altogether, we demonstrate that diabetes induces M1 but decreases M2 cytokine expression in adipose tissue. Diabetic plasma-induced M1 expression is potentially through pJNK signaling pathways. Non-diabetic plasma reverses M1/M2 cytokine expression, plasma CCL2 levels, DPP4 activity, and Kupffer cell activation in diabetes. Our results suggest M1/M2 cytokine expression in adipose tissue is critical in diabetes-induced DPP4 activity, liver inflammation, and insulin resistance.


Introduction
Dysregulation of M1/M2 polarization in adipose tissue is one of the major mechanisms underlying the pathogenesis of obesity and comorbidities, such as insulin resistance and nonalcoholic fatty liver disease [1]. In cases of obesity, adipose tissue macrophages change from the anti-inflammatory M2 phenotype to the pro-inflammatory M1 phenotype. These M1 macrophages produce inflammatory cytokines, such as TNF-α, IL-6, and IL-1β, that inhibit the ability of adipocytes to respond to insulin. M1 macrophages secrete cytokines, such as TNF-α and IL-6, to activate inflammatory pathways in insulin target cells, resulting in the activation of Jun N-terminal kinase (JNK) and inhibition of κB kinase β [1]. M2 adipose tissue macrophages, which are the major resident macrophages in lean adipose tissue, are characterized by high expression of IL-10 and CD206 [2]. The depletion of M1 macrophages normalizes sensitivity to insulin in obese mice, whereas the reduction of M2 macrophage numbers predisposes lean subjects to insulin resistance [2]. Indeed, studies show there is a close correlation between adipose tissue inflammation and metabolic diseases in obese humans [3,4]. Chronic inflammation is one of the major causes of vascular and kidney complications in patients with diabetes [5]. The inflammatory response could be stimulated by various mechanisms, including hyperglycemia-induced cell death, which increases the aggregation of macrophages in the kidney [6]. Blocking systemic or adipose tissue macrophage inflammation has been shown to improve insulin sensitivity in obese mice [7]. However, the regulatory mechanisms of M1/M2 polarization in adipose tissue under chronic inflammation and related complications in diabetes have not been well characterized.
Dipeptidyl peptidase 4 (DPP4) is a complex enzyme that is present on the surface of many cell types, including kidney, liver, pancreas, and fat cells, and also present as a soluble form in the circulation [8]. DPP4, a serine protease, cleaves the penultimate L-proline or L-alanine found at the N-terminal of several polypeptides, such as glucagon-like peptide 1 (GLP-1), glucose-dependent insulinotropic polypeptide (GIP), neuropeptides, and chemokines [9]. These peptides are rapidly degraded and inactivated by DPP4. In addition, DPP4 has been shown to exert a direct pro-inflammatory role in different cell types, including lymphocytes, macrophages, and smooth muscle cells [10,11]. Hematopoietic cells have been found to be a major source of circulating soluble DPP-4 (sDPP4). Moreover, activation of mouse and human T lymphocytes induced sDPP4 [12]. Previously, a study showed that plasma DPP4 levels and DPP4 mRNA expression in tissues increased progressively during the development of streptozotocin-induced type 1 diabetes [13]. Moreover, DPP4 expression in inflammatory cells such as dendritic cells and macrophages may have a significant role in modulating the adipose tissue inflammation in obesity through its nonenzymatic function [14]. Recently, a study reported that the circulating levels of endogenous soluble DPP4 were dissociated from the extent of systemic and white adipose tissue inflammation [15]. The study demonstrated that knockdown of adipocyte DPP4 expression decreased plasma DPP4 activity in older mice but did not change incretin levels or glucose homeostasis. In addition, the study found that the genetic disruption of hepatocyte DPP4 resulted in abrogation of obesity-associated increase of plasma DPP4 activity, repression of liver cytokine expression, and partly amelioration of inflammation in adipose tissue, however the incretin levels and glucose homeostasis were not affected [15]. These observations revealed complex biological processes in the regulation of soluble and enzymatic DPP4 as well as in the control of inflammation and glucose homeostasis.
In this study, we examined the involvement of adipose tissue M1/M2 cytokine expression in plasma DPP4 activity, systemic inflammation, and insulin resistance in a diabetic mouse model. We hypothesize that diabetes induces M1 but decreases M2 cytokine expression in adipose tissue. Increased M1 expression induces DPP4 activity in plasma and upregulates inflammatory mediators in liver that subsequently lead to systemic inflammation and insulin resistance in diabetes. Based on that, we first determined the changes of M1/M2 cytokine expression in adipose tissue of diabetic and non-diabetic mice. Furthermore, we explored whether diabetic plasma increases M1 expression and non-diabetic plasma decreases M1 expression in SVFs from adipose tissue of diabetes. Finally, we examined whether the injection of non-diabetic plasma reverses M1 cytokine expression, plasma DPP4 activity and CCL2 levels, systemic inflammation, and glucose intolerance in diabetes. Our findings provide a new insight that the inhibition of M1 cytokine expression in adipose tissue may represent a novel therapeutic approach for attenuating DPP4 activity, systemic inflammation, and insulin resistance in diabetic patients. Our findings provide a new insight that the inhibition of M1 cytokine expression in adipose tissue with non-diabetic plasma may represent a novel therapeutic approach for attenuating DPP4 activity, systemic inflammation, and insulin resistance in diabetic patients.

Animals and treatments
Lepr db/+ mice on the C57BL/6J genetic background were purchased from the Jackson Laboratory (Bar Harbor, ME). Lepr db/+ mice were bred to obtain diabetic Lepr db/db (total n = 120) and non-diabetic Lepr +/+ mice (total n = 80). Lepr db/db mice that have a mutation in the gene encoding the leptin receptor become obese at 3 to 4 weeks of age, and elevated plasma insulin and blood sugar can be observed in Lepr db/db mice at four to eight weeks of age. Jnk1 -/-(c-Jun N-terminal kinases 1 knockout) mice on the C57BL/6J background were transferred from Dr. Karin's laboratory (University of California, San Diego, CA, USA). Jnk1 -/mice were crossbred with Lepr db/+ mice to obtain Lepr db/+ -Jnk1 -/mice, and Lepr db/+ -Jnk1 -/mice were then bred to obtain Lepr db/db -Jnk1 -/mice (total n = 30). All mice had ad libitum access to water and food and were fed a standard laboratory diet (1324 TPF; Atromin; Large Germany; 11.9 kJ/g, 19% crude protein, 4% crude fat, 6% crude fiber).

Ethics statement
The Institutional Animal Care and Use Committee (IACUC) of Kaohsiung Veterans General Hospital has approved this study (Approval Number: 2020-A042), and animal experiments were performed in agreement with Animal Experimentation Regulations of Kaohsiung Veterans General Hospital. Animals were observed every 12 hours. All efforts were made to reduce suffering of animals. Personnel conducting mice treatment received training and competency testing from Kaohsiung Veterans General Hospital. Mice were euthanized with carbon dioxide when they were found in a moribund state as identified by inability to maintain upright with or without labored breathing and cyanosis, >15% loss of body weight, inability to obtain food or water, or unresponsive to normal physical manipulation. Once any animal reached endpoint criteria, the amount of time elapsed before euthanasia was less than 12 hours. No mortality occurred outside of planned euthanasia or humane endpoints.

Preparation of SVFs
SVFs were prepared from Lepr db/db or Lepr +/+ mice. Vascular adipose tissue was isolated from bilateral inguinal adipose tissue of diabetic and non-diabetic mice and minced into small pieces. These tissues were then digested with collagenase 8 (Sigma-Aldrich, Cat# C2139) in ice-cold HBSS (2 mg/ml) for 15 minutes at 37˚C. After passing through a 100 μm cell strainer, cells were centrifuged at 1,200 rpm for 10 min, and the cell pellets were retrieved as SVFs.

In vitro treatment of SVFs
For in vitro treatment, 1ml of 10% plasma was added to the equal amount of SVFs (1 × 10 9 cells/ml), and the mixture was incubated at 37˚C for 4 hr as previously suggested [11]. Samples were then centrifuged at 2,500 rpm for 5 min and washed with phosphate buffered saline (PBS). After centrifugation, the pellets were harvested and subjected to analysis.

In vivo injection of plasma into inguinal adipose tissue
Male Lepr db/db mice were randomly divided into the following three groups: Group I received 1ml PBS injection into adipose tissue over the bilateral inguinal area; Group II received an injection of 1ml 10% plasma from Lepr db/db mice into adipose tissue over the bilateral inguinal area; and Group III received an injection of 1ml 10% plasma from Lepr +/+ mice into adipose tissue over the bilateral inguinal area. Seven days after injection, the animals were euthanized and liver, adipose tissue, and plasma were harvested for analysis.

RNA isolation and quantitative real-time polymerase chain reaction (Q-PCR)
Total RNA was purified from mouse samples using total RNA Miniprep Purification Kits (GeneMark) according to the manufacturer's instructions. Total RNA was then reverse-transcribed into cDNA using a RT kit (Invitrogen, Carlsbad, CA, Lot# 2234812). Primer pairs were synthesized by Integrated DNA Technologies (Coralville, IA). Primer sequences are listed in S1 Table. For Q-PCR assay, 200 ng of the cDNA template was added to 25 μl of mixture containing 12.5 μl of 2× Fast SYBR Green Master Mix (Applied biosystems, Cat# 4385612), 2.5 μl of sense and antisense primers (25 μM), 2 μl of sample, and 8 μl of sterile water. The amplification was performed in a StepOnePlus™ Real-Time PCR System (Applied Biosystems 7300).

Western immunoblots
The protein expression of DPP4 (GeneTex, Cat# GTX84602), phospho-Akt (Cell signaling, Cat# 4060), Akt (Cell signaling, Cat# 4691), phospho-p38 (Cell signaling, Cat# 9211), p38 (Cell signaling, Cat# 9212), FMO3 (Novus, Cat# NBP1-33583), ICAM-1 (Abnova, Cat# E5081) in the liver as well as JNK (Cell signaling, Cat# 9252) and pJNK (Cell signaling, Cat# 9251) in the SVFs was detected by antibodies. The harvested tissues were weighed and homogenized in protein extraction buffer (Sigma), containing proteinase inhibitor cocktail (Roche). The homogenized samples were subjected to SDS-PAGE at 50 to 100V for 2 hours, and the proteins were then transferred onto the nitrocellulose membrane. The membranes were blocked with 5% non-fat milk in TBST buffer (10 mM Tris-HCl, pH 7.5, 150 mM NaCl and 1.2% Tween 20) for 1 hour and incubated with specific primary antibodies at room temperature for 1 hour. The membranes were then washed with TBST buffer followed by incubation with the secondary antibodies. After washing with TBST buffer, the protein bands were identified by enhanced chemiluminescence (ECL) detection reagent (Millipore).

Kupffer cell purification
The liver was perfused in situ through the portal vein with Ca 2+ -and Mg 2+ -free phosphatebuffered saline containing 10 mM ethylenediaminetetraacetic acid at 37˚C for 5 minutes. Subsequently, perfusion was performed with HBSS containing 0.1% collagenase IV (Sigma) at 37˚C for 5 minutes. After digestion, the liver was excised and the suspension was filtered. The sample was centrifuged twice at 50g at 4˚C for 1 minute. The supernatant was harvested and centrifuged at 300g for 5 minutes, and the pellet was collected and resuspended with buffer. The cell suspension was then layered on top of a density cushion of 30/60 discontinuous Percoll (Pharmacia) and centrifuged at 900g for 15 minutes to obtain the Kupffer cell fraction [16].

Enzyme-linked immunosorbent assay (ELISA)
The mouse ELISA kit (R&D, Minneapolis, MN, Cat# DY479-05) was used for CCL2 assay. The blood was centrifuged at 1000×g, 4˚C for 15 minutes and the serum was then collected for use. The ELISA plates were coated with 100 μl capture antibody at 4˚C overnight. After wash, 200 μl of assay dilution buffer were added for blocking at room temperature for 1 hour. The samples and serial dilutions of standards were added and incubated at 4˚C for overnight. After incubating with the detection antibody, avidin-HRP was added and incubated at room temperature for 30 minutes. The substrate, 3,3 0 ,5,5 0 -tetramethylbenzidine (TMB), was then added and incubated for 15 minutes. 100 μl of stop solution were added to stop reaction, and the plate was then subjected to measurement of absorbance at 450 nm by using an ELISA reader.

Intraperitoneal glucose tolerance tests (IPGTTs)
Glucose tolerance was assessed using intraperitoneal glucose tolerance tests (IPGTTs). Mice were tested in the morning after a 15-hour fasting period. Tail blood glucose was measured before and at 5, 15, 30, 45, 60, 75, 90, and 120 minutes after intraperitoneal injection of glucose (2 g/kg body weight; Sigma). Plasma glucose was measured using a glucose meter (Accu-check performa; Roche, Switzerland).

Statistical analysis
Data were analyzed by unpaired t test for the comparisons between two groups or by one-way analysis of variance (ANOVA) followed by Tukey's Multiple comparison test for the comparisons between multiple groups. All values in the figures and texts were expressed as mean ± standard error of the mean, and p values less than 0.05 are considered statistically significant.

SVFs from Lepr db/db mice exhibit increased M1 expression and SVFs from Lepr +/+ mice exhibit increased M2 expression
To study M1/M2 polarization in the adipose tissue of diabetes and non-diabetes, we examined M1/M2 cytokine mRNA expression in SVFs of inguinal adipose tissue from type 2 diabetic (Lepr db/db ) mice and control non-diabetic (Lepr +/+ ) mice. SVFs from Lepr db/db mice demonstrated a significant induction of TNF-α, IL-6, IL-1β, CCL2, and DPP4 mRNA expression compared with those from Lepr +/+ mice. In contrast, SVFs from Lepr db/db mice exhibited a significant reduction of IL-10 compared with those from Lepr +/+ mice (Fig 1). Collectively, these results indicate that diabetes induces M1 but reduces M2 cytokine expression in the adipose tissue.

PLOS ONE
Inhibiting adipose M1 cytokines decreases DPP4 and insulin resistance

Diabetic plasma induces M1 expression and non-diabetic plasma induces M2 expression in SVFs of Lepr db/db mice
To examine the effects of diabetic and non-diabetic plasma on M1/M2 polarization in the adipose tissue of diabetes, SVFs were harvested from the adipose tissue of Lepr db/db mice and treated with plasma from Lepr db/db or Lepr +/+ mice in vitro. Treatment of Lepr db/db SVFs with Lepr db/db plasma demonstrated a significant upregulation of TNF-α, IL-6, IL-1β, CCL2, and DPP4 mRNA expression compared with those treated with PBS (Fig 2). On the contrary, treatment of Lepr db/db SVFs with Lepr +/+ plasma resulted in a significant downregulation of aforementioned inflammatory mediators and upregulation of IL-10 mRNA expression compared with those treated with Lepr db/db plasma (Fig 2). Altogether, these results demonstrate that diabetic plasma induces M1 expression, however non-diabetic plasma induces M2 expression in the adipose tissue of diabetic mice.

Injection of plasma from Lepr +/+ mice into adipose tissue of Lepr db/db mice reduces M1 expression in vivo
To further evaluate the effects of diabetic and non-diabetic plasma on M1/M2 polarization of adipose tissue in diabetic mice, plasma harvested from Lepr db/db and Lepr +/+ mice was injected into the adipose tissue near bilateral inguinal area in Lepr db/db mice, and SVFs were then isolated from the adipose tissue followed by Q-PCR analysis to determine cytokine expression. Our results showed that Lepr db/db mice injected with PBS exhibited increased expression of TNF-α, IL-6, IL-1β, CCL2, and DPP4 mRNA expression in SVFs compared to Lepr +/+ mice injected with PBS ( Fig 3A). Notably, injection of Lepr +/+ plasma significantly induced IL-10 but decreased TNF-α, IL-6, IL-1β, and DPP4 mRNA expression in SVFs of Lepr db/db mice compared with those injected with Lepr db/db plasma (Fig 3A). Taken altogether, these results demonstrate that non-diabetic plasma reduces M1 cytokine but increases M2 cytokine expression in diabetic adipose tissue.

Injection of plasma from Lepr +/+ mice into adipose tissue of Lepr db/db mice decreases JNK activation and DPP4 protein expression in adipose tissue of Lepr db/db mice
To investigate the effects of diabetic and non-diabetic plasma on JNK activation and DPP4 expression in the adipose tissue of diabetes, plasma harvested from Lepr db/db and Lepr +/+ mice was injected into the adipose tissue of Lepr db/db mice, and JNK activation and DPP4 protein expression in SVFs of adipose tissue were determined. SVFs from Lepr db/db mice displayed increased JNK activation and DPP4 protein expression compared with those from Lepr +/+ mice (Fig 3B). In contrast, injection of Lepr +/+ plasma mice into the adipose tissue of Lepr db/db mice significantly decreased JNK activation and DPP4 protein expression in SVFs of Lepr db/db mice compared with those receiving Lepr db/db plasma. Altogether, these results suggest that diabetes induces JNK activation and DPP4 protein expression and non-diabetic plasma diminishes diabetes-induced JNK activation and DPP4 protein expression in the adipose tissue of diabetes.

Injection of plasma from Lepr +/+ mice into adipose tissue of Lepr db/db mice decreases inflammatory cytokine expression in the liver of Lepr db/db mice
To further explore the effects of diabetic and non-diabetic plasma on cytokine expression in the liver of diabetes, plasma harvested from Lepr db/db and Lepr +/+ mice was injected into the adipose tissue of Lepr db/db mice, and the cytokine expression in the liver of Lepr db/db mice was  evaluated. Lepr db/db mice exhibited increased mRNA expression of ICAM, FMO3, IL-1β, iNOS, TNF-α, IL-6, and DPP4 in the liver compared with Lepr +/+ mice (Fig 4A). Notably, injection of Lepr +/+ plasma into the adipose tissue of Lepr db/db mice largely decreased mRNA expression of ICAM, FMO3, IL-1β, iNOS, TNF-α, IL-6, and DPP4 in the liver of Lepr db/db mice compared with those receiving the injection of Lepr db/db plasma (Fig 4A). Altogether, these results demonstrate that non-diabetic plasma suppresses the expression of inflammatory mediators in the liver of diabetic mice.
Injection of plasma from Lepr +/+ mice decreases ICAM, FMO3, and DPP4 protein expression in the liver of Lepr db/db mice To elucidate the effects of diabetic and non-diabetic plasma on ICAM, FMO3, and DPP4 protein expression in the liver of diabetes, plasma from Lepr db/db and Lepr +/+ mice was injected into the adipose tissue of Lepr db/db mice, and the protein expression of ICAM, DPP4, and FMO3 in the liver was examined. Lepr db/db mice displayed increased liver ICAM, FMO3, and DPP4 protein expression compared with Lepr +/+ mice (Fig 4B). Injection of Lepr +/+ plasma into the adipose tissue significantly decreased ICAM, FMO3, and DPP4 protein expression in the liver of Lepr db/db mice compared with those receiving Lepr db/db plasma. Collectively, these results indicate that diabetes induces ICAM, FMO3, and DPP4 expression in the liver, and non-diabetic plasma injected into the adipose tissue reduces the expression of those proteins in the liver of diabetic mice.

Injection of plasma from Lepr +/+ mice to adipose tissue of Lepr db/db mice decreases TNF-α, IL-6, INOS, and DPP4 mRNA expression of Kupffer cells in Lepr db/db mice
To further examine the effects of diabetic and non-diabetic plasma on inflammatory cytokine expression in Kupffer cells of diabetes, plasma harvested from Lepr db/db and Lepr +/+ mice was injected into the adipose tissue of Lepr db/db mice, and Kupffer cells were then isolated from the liver of Lepr db/db mice and subjected to Q-PCR analysis to determine mRNA expression of inflammatory cytokines. Kupffer cells isolated from Lepr db/db mice displayed increased mRNA expression of TNF-α, INOS, DPP4, and IL-6 compared with those isolated from Lepr +/+ mice (Fig 5). In contrast, Kupffer cells isolated from Lepr db/db mice treated with Lepr +/+ plasma showed repressed mRNA expression of aforementioned inflammatory mediators compared with those isolated from Lepr db/db mice treated with Lepr db/db plasma (Fig 5). Altogether, these results demonstrate that injection of non-diabetic plasma into the adipose tissue is capable of suppressing inflammatory mediator production in Kupffer cells of diabetic mice.

Lepr db/db mice exhibit increased plasma DPP4 activity, whereas injection of Lepr +/+ plasma decreases plasma DPP4 levels in Lepr db/db mice
To examine whether there is a difference in plasma DPP4 activity between diabetic and nondiabetic mice, we measured plasma DPP4 activity in Lepr db/db and Lepr +/+ mice of both sexes. We observed that Lepr db/db mice exhibited higher plasma DPP4 activity compared with Lepr +/+ mice in both males and females (Fig 6A). We then further determine the effects of diabetic and non-diabetic plasma on plasma DPP4 activity in diabetes. Plasma harvested from Lepr db/db or Lepr +/+ mice was injected into the adipose tissue of Lepr db/db mice, and plasma DPP4 activity was then assessed. Our results showed that injection of Lepr +/+ plasma into the adipose tissue of Lepr db/db mice resulted in a significant reduction of plasma DPP4 activity in Lepr db/db mice compared with those receiving Lepr db/db plasma (145.5 ± 11.61 vs. 199 ± 7.1 pmol/min/ml×10 −3 )

Fig 4. Injection of plasma from Lepr +/+ mice into adipose tissue of Lepr db/db mice decreases inflammatory cytokine expression as well as ICAM, FMO3, and DPP4 protein expression in the liver of Lepr db/db mice. Plasma harvested from
Lepr db/db or Lepr +/+ mice was diluted with PBS to 10%. One ml of 10% plasma was then injected to adipose tissue of Lepr db/db mice. Lepr db/db and Lepr +/+ mice were injected with PBS to serve as controls. Seven days after injection, animals were ( Fig 6B). Collectively, these results demonstrate that Lepr db/db mice exhibit increased plasma DPP4 activity, and non-diabetic plasma is capable of repressing plasma DPP4 activity in Lepr db/db mice.

Fig 5. Injection of plasma from Lepr +/+ mice into adipose tissue of Lepr db/db mice decreases TNF-α, IL-6, iNOS, and DPP4 mRNA expression in Kupffer cells of Lepr db/db mice.
Plasma harvested from Lepr db/db or Lepr +/+ mice was diluted with PBS to 10%, and one ml of 10% plasma was injected into adipose tissue of Lepr db/db mice. Lepr db/db and Lepr +/+ mice were also injected with PBS to serve as controls. Seven days after injection, animals were euthanized and livers were harvested. Kupffer cells were then isolated from livers and subjected to Q-PCR analysis for mRNA expression of TNFα, iNOS, DPP4, and IL-6. N = 6/group. � p<0.05; �� p<0.01; ��� p<0.001.

Injection of plasma from Lepr +/+ mice increases Akt activation in Lepr db/db mice following insulin administration
To investigate whether non-diabetic plasma affects insulin resistance in diabetes, Lepr db/db mice were subjected to Lepr +/+ plasma injection, and Lepr +/+ , Lepr db/db , and Lepr +/+ plasmainjected Lepr db/db mice were administered with or without insulin to determine the level of Akt phosphorylation in the liver. Akt phosphorylation was only observed in the liver of Lepr +/+ mice but not Lepr db/db mice following insulin administration, indicating that Lepr db/db mice did not respond to insulin treatment (Fig 7A and 7B). However, Akt phosphorylation was able to be observed in the liver of Lepr db/db mice injected with Lepr +/+ plasma following insulin administration (Fig 7A and 7B), suggesting that Lepr +/+ plasma promotes insulin response in Lepr db/db mice. In addition, Lepr db/db mice exhibited increased DPP4 expression in the liver compared to Lepr +/+ mice, and the injection of Lepr +/+ plasma was able to suppress DPP4 expression in the liver of Lepr db/db mice. Collectively, these results indicate that diabetes induces DPP4 expression and suppress insulin-induced Akt activation in the liver, and the injection of non-diabetic plasma into the adipose tissue of diabetic animals reverses those effects.

Injection of plasma from Lepr +/+ mice reverses glucose intolerance in Lepr db/db mice
To further examine the effect of injection of Lepr +/+ plasma on diabetes-induced glucose intolerance, we performed glucose tolerance test in Lepr +/+ , Lepr db/db , and Lepr +/+ plasma-injected Lepr db/db mice. After fasting, Lepr +/+ mice showed an increased blood glucose level after glucose administration, however, the blood glucose level returned to the normal level at 2 hours post administration. In contrast, Lepr db/db mice maintained a high blood glucose level even at 2 hours after glucose administration. Notably, Lepr +/+ plasma-treated Lepr db/db mice, unlike Lepr db/db mice, exhibited a reduced blood glucose level post glucose administration (Fig 7C). Altogether, these results demonstrate that diabetes induces glucose intolerance, however nondiabetic plasma ameliorates impaired glucose tolerance in diabetic mice.

Diabetic plasma induces M1 expression in adipose tissue from Lepr db/db mice but not in adipose tissue from Lepr db/db -Jnk1 -/mice
To examine whether JNK activation plays an essential role in adipose tissue M1 expression in diabetes, SVFs harvested from adipose tissue of Lepr db/db and Lepr db/db -Jnk1 -/mice were treated with PBS or Lepr db/db plasma in vitro. Lepr db/db SVFs treated with Lepr db/db plasma demonstrated increased mRNA expression of TNF-α, IL-6, IL-1β, CCL2, and DPP4 compared with Lepr db/db SVFs treated with PBS (Fig 8). In contrast, Lepr db/db -Jnk1 -/-SVFs treated with Lepr db/db plasma did not exhibit increased mRNA expression of aforementioned inflammatory mediators compared with Lepr db/db -Jnk1 -/-SVFs treated with PBS (Fig 8). Altogether, these results suggest that JNK activation may play an essential role in promoting adipose tissue M1 polarization in diabetic mice. assay kit. (B and C) Lepr db/db mice were injected with 1ml of 10% plasma harvested from Lepr db/db or Lepr +/+ mice. Lepr db/db and Lepr +/+ mice were also injected with PBS to serve as controls. After seven days, animals were euthanized and plasma was harvested to detect DPP4 activity and CCL2 level. N = 5/group. � p<0.05; �� p<0.01; ��� p<0.001. https://doi.org/10.1371/journal.pone.0252153.g006

Discussion
Adipose tissue dysfunction is characterized by low-grade chronic inflammation in diabetes, metabolic disorders and insulin resistance [17]. Chronic inflammation is one of the major causes for vascular and kidney complications in patients with diabetes. In this study, we demonstrated that M1/M2 cytokine expression in adipose tissue may play an important role in diabetes-induced systemic inflammation and importantly, the reversal of diabetes-induced M1/M2 cytokine expression in the adipose tissue my represent a novel therapeutic strategy for attenuating chronic inflammation-induced systemic complications in diabetes.
M1 macrophages secrete cytokines to activate inflammatory pathways in insulin-target cells, which results in the activation of JNK, κB kinase β inhibitors, and other serine kinases [1]. Our data demonstrated that SVFs from adipose tissue of Lepr db/db mice exhibited enhanced JNK activation compared with those from Lepr +/+ mice and injection of plasma from Lepr +/+ mice into adipose tissue of Lepr db/db mice suppressed JNK activation compared with those injected with plasma from Lepr db/db mice. Importantly, plasma from Lepr db/db mice induced TNF-α, IL-6, IL-1β, CCL2, and DPP4 mRNA expression in SVFs from Lepr db/db mice but not in SVFs from Lepr db/db -Jnk1 -/mice. These results suggest that M1 cytokines activate JNK pathway to promote inflammatory cytokine production and injection of non-diabetic plasma may suppress JNK activation and inflammatory cytokine production in the adipose tissue of diabetes.
DPP4 activity is associated with the onset and severity of obesity and diabetes [13]. The levels of plasma DPP4 activity are elevated in diseases, including type 2 diabetes mellitus (T2DM) [18], obesity [18], and atherosclerosis [19]. Recently, the circulating levels of endogenous soluble DPP4 were found to be dissociated from the extent of systemic and white adipose tissue inflammation [15]. However, we found that diabetes-induced plasma DPP4 activity and DPP4 expression in tissues were closely related with diabetes-induced systemic inflammation and glucose intolerance. First, both female and male Lepr db/db mice demonstrated increased plasma DPP4 activity compared with their corresponding Lepr +/+ mice. Second, Lepr db/db mice exhibited increased DPP4 mRNA expression in adipose tissue and liver compared with Lepr +/+ mice. Third, in vitro treatment of Lepr db/db SVFs with diabetic plasma significantly induced DPP4 mRNA expression compared with those treated with non-diabetic plasma. However, injection of non-diabetic plasma to adipose tissue of diabetic mice resulted in decreased plasma DPP4 activity and repressed DPP4 expression in adipose tissue and liver in diabetic mice compared to those injected with diabetic plasma. Importantly, injection of non-diabetic plasma into the adipose tissue of diabetic mice attenuated glucose intolerance and insulin resistance in diabetic mice. Altogether, our results demonstrate that diabetes induces adipose tissue and liver DPP4 expression as well as plasma DPP4 activity. In addition, our findings further reveal that the reversal of DPP4 expression in adipose tissue with non-diabetic plasma not only decreases plasma DPP4 activity and liver DPP4 expression but also ameliorates glucose intolerance and insulin resistance in diabetes. Thus our findings suggest that the inhibition of diabetes-induced DPP4 activity could be a plausible therapeutic strategy to reestablish glucose homeostasis in diabetes.
The increased secretions of multiple proinflammatory cytokines and chemokines [20], such as IL-6, TNF-α, IL-1β, and monocyte chemoattractant protein-1 (MCP-1) in adipose tissue are involved in insulin resistance [21]. Ghorpade et al. demonstrated that 10% (vol/vol) plasma from diet-induced obesity mice (DIO) induced higher expression of both MCP1 and IL-6 mRNA expression than 10% plasma from lean mice in SVFs from DIO mice [11]. Furthermore, administration of adipose tissue macrophages from lean mice to obese recipients has been proved to improve glucose tolerance and insulin sensitivity in obese mice through the secretion of miRNA-containing exosomes [22]. Our data demonstrated that 10% diabetic plasma induced M1 expression, however 10% non-diabetic plasma induced M2 expression in SVFs from diabetic mice. Moreover, injection of non-diabetic plasma reduces M1 cytokine but increases M2 cytokine expression in diabetic adipose tissue. Our data demonstrated that the injection of non-diabetic plasma could be a promising therapeutic strategy to reverse diabetes-induced systemic inflammation and insulin resistance through inhibiting inflammatory cytokines. First, in vitro treatment of Lepr db/db SVFs with Lepr +/+ plasma decreased TNF-α, IL-6, IL-1β, CCL2, and DPP4 but increased IL-10 mRNA expression. Second, injection of nondiabetic plasma into adipose tissue of Lepr db/db mice decreased TNF-α, IL-6, IL-1β, CCL2, and DPP4 mRNA expression as well as pJNK and DPP4 protein expression while increasing IL-10 mRNA expression in adipose tissue of diabetic mice.
Third, injection of non-diabetic plasma decreased ICAM, FMO3, IL-1β, iNOS, TNF-α, and IL-6 mRNA expression and ICAM and FMO3 protein expression in liver, and suppressed TNF-α, iNOS, and DPP4 mRNA expression in Kupffer cells in diabetic animals. Previously, we have demonstrated that diabetes increases intestinal iNOS expression, NO levels in the portal vein, IL-1β and TNF-α expression of Kupffer cells in the liver. iNOS inhibition by L-NAME reduces NO levels in the portal vein, IL-1β and TNF-α expression of Kupffer cells in the liver [23]. Here, we demonstrated that injection of Lepr +/+ plasma into the adipose tissue of Lepr db/db mice decreased mRNA expression of proinflammatory cytokines and DPP4 in Kupffer cells of diabetic mice. Finally, injection of non-diabetic plasma decreased plasma DPP4 activity and CCL2 levels, and ameliorated insulin resistance. Taken altogether, these results suggest that non-diabetic plasma injection into diabetic adipose tissue could be used to reverse diabetes-induced systemic inflammation and insulin resistance through the reversal of M1/M2 cytokine expression.
M2 macrophages are thought to facilitate adipose tissue homeostasis and protect against insulin resistance. M2 polarization of adipose tissue macrophages can be promoted by Th2 cytokine stimulation, such as IL-4 and IL-13. IL-6, a classical proinflammatory cytokine induced during different kinds of inflammation was shown to provoke M2 polarization of adipose tissue [24]. Furthermore, studies also identify multiple adipose tissue-specific mediators of M2 polarization, including transcription factors, adipokines, fatty acids, and other immune cells [25]. Nevertheless, studies have suggested that the M1/M2 paradigm is an oversimplification of the macrophage biology observed in vivo [26]. Thus, the term "metabolic activation" has been suggested by Kratz et al. [26], which is characterized by an increased machinery of lipid catabolism in adipose tissue macrophages. Our data suggest that non-diabetic plasma injection into adipose tissue could be used to reverse diabetes-induced systemic inflammation and insulin resistance through the reversal of M1/M2 cytokine expression. We have examined different pro-inflammatory cytokine levels in the plasma injected mice. Although injection of Lepr +/+ plasma into the adipose tissue of Lepr db/db mice did not change plasma levels of TNF-α and IL-6 compared with those injected with Lepr db/db plasma, injection of Lepr +/+ plasma significantly decreased plasma levels of CCL2 in diabetic mice compared with those injected with Lepr db/db plasma. On the other hand, treatment of Lepr db/db plasma induced upregulation of inflammatory mediators, TNF-α, IL-6, IL-1β, CCL2, and DPP4, in Lepr db/db SVFs. However, that effect was not observed in Lepr db/db -Jnk1 -/-SVFs treated with Lepr db/db plasma. Thus, we speculate that diabetic plasma induces JNK activation to promote M1 polarization of adipose tissue in diabetes. Whether non-diabetic plasma regulates JNK activation to promote M2 polarization in diabetic adipose tissue would require further investigations.
Our study had several limitations. We used a single injection and found out that the single injection of non-diabetic plasma could reverse M1/M2 cytokine expression, DPP4 activity, inflammatory cytokine expression of liver, and Kupffer cell activation in diabetes. Future studies using multiple injections may have more prominent effect. We did not examine M1/M2 cytokine production in peritoneal macrophages. Also, our data suggest that injection of non-diabetic plasma into adipose tissue could be used to reverse diabetes-induced systemic inflammation and insulin resistance through the reversal of M1/M2 cytokine expression. Whether non-diabetic plasma regulates JNK activation to promote M2 polarization in diabetic adipose tissue would require further investigations.
In summary, we demonstrate the involvement of M1/M2 cytokine expression in diabetesinduced DPP4 activity, insulin resistance, and systemic inflammation (Fig 9). Diabetes induces M1 cytokine expression and decreases M2 cytokine expression in diabetic adipose tissue. Increased adipose tissue M1 expression induces JNK activation, plasma DPP4 activity, and liver ICAM, FMO3, IL-1β, iNOS, TNF-α, IL-6, and DPP4 expression. Subsequently, increased inflammatory cytokine expression and plasma DPP4 activity may promote insulin resistance and glucose intolerance in diabetes. On the other hand, non-diabetic plasma inhibits M1 expression and increases M2 expression in the adipose tissue of diabetes. The inhibition of M1 expression in the adipose tissue represses JNK activation, plasma DPP4 activity and CCL2 levels, as well as liver inflammatory cytokines expression that may ultimately attenuate insulin resistance and glucose intolerance in diabetes.
Supporting information S1 Checklist. (DOCX) Fig 9. The model of non-diabetic plasma-mediated suppression of M1 expression and attenuation of insulin resistance and glucose intolerance. M1 expression in adipose tissue promotes adipose JNK activation, plasma DPP4 and CCL2 induction, and liver ICAM, FMO3, IL-1β, iNOS, and DPP4 upregulation that may result in insulin resistance and impaired glucose tolerance in diabetes. On the other hand, non-diabetic plasma represses diabeticinduced M1 expression in adipose tissue, plasma, and liver that may ultimately attenuate insulin resistance.