Repeated clodronate-liposome treatment results in neutrophilia and is not effective in limiting obesity-linked metabolic impairments

Abstract

Depletion of macrophages is thought to be a therapeutic option for obesity-induced inflammation and metabolic dysfunction. However, whether the therapeutic effect is a direct result of reduced macrophage-derived inflammation or secondary to decreases in fat mass is controversial, as macrophage depletion has been shown to disrupt energy homeostasis. This study was designed to determine if macrophage depletion via clodronate-liposome (CLD) treatment could serve as an effective intervention to reduce obesity-driven inflammatory and metabolic impairments independent of changes in energy intake. After 16 wk on a high-fat diet (HFD) or the AIN-76A control (low-fat) diet (LFD) (n = 30/diet treatment), male C57BL/6J mice were assigned to a CLD- or PBS-liposome treatment (n = 15/group) for 4 wk. Liposomes were administered biweekly via intraperitoneal injections (8 administrations in total). PBS-liposome-treated groups were pair-fed to their CLD-treated dietary counterparts. Metabolic function was assessed before and after liposome treatment. Adipose tissue, as well as the liver, was investigated for macrophage infiltration and the presence of inflammatory mediators. Additionally, a complete blood count was performed. CLD treatment reduced energy intake. When controlling for energy intake, CLD treatment was unable to regress metabolic dysfunction or nonalcoholic fatty liver disease and impaired adipose tissue insulin action. Moreover, repeated CLD treatment induced neutrophilia and anemia, increased adipose tissue mRNA expression of the proinflammatory cytokines IL-6 and IL-1β, and augmented circulating IL-6 and monocyte chemoattractant protein-1 concentrations (P < 0.05). This study suggests that repeated intraperitoneal administration of CLD to deplete macrophages attenuates obesity by limiting energy intake. Moreover, after controlling for the benefits of weight loss, the accompanying detrimental side effects limit regular CLD treatment as an effective therapeutic strategy.

INTRODUCTION

Obesity has been established as a public health concern, given the condition’s link to increased health care costs, morbidity, and mortality (34). The underlying cause of obesity is rooted in a lack of physical activity and an excessive intake of energy-dense foods, resulting in the obese phenotype. This phenotype is largely defined by two primary characteristics: chronic, low-grade inflammation and metabolic dysfunction. Although it has been established that these two processes are linked, the degree to which one process regulates the other is not completely understood. It is believed that a main driving force behind the development of type 2 diabetes and insulin resistance is inflammation (5). Thus, therapies designed to combat obesity-associated inflammation may protect against the obesity-associated metabolic diseases.

Many different factors, including, but not limited to, changes in the gut microbiota, the gut-brain axis and nonalcoholic fatty liver disease (NAFLD), are thought to play a role in obesity-associated inflammation (4, 38). However, the immune cells primarily responsible for adipose tissue inflammation are infiltrating proinflammatory (M1) macrophages (45). Under obese conditions, macrophages, predominantly of the M1 phenotype, may account for ~50% of all adipose tissue cells (45). This is in stark contrast to the lean condition, in which macrophages account for only ~10% of all adipose tissue cells, with the anti-inflammatory (M2) macrophage being the most predominant. In addition to clearing dead adipocytes, M1 macrophages secrete a variety of proinflammatory cytokines, including TNF-α, a cytokine known to negatively impact insulin signaling (22). As such, depletion of macrophages to reduce macrophage-derived adipose tissue inflammation has been investigated as a potential therapeutic strategy. Depletion of macrophages in an obese setting reduces adipose tissue inflammation and improves insulin sensitivity (6, 17). Macrophage depletion has also been shown to reduce energy intake (29) and body weight (20, 25, 29, 47), and it is well known that energy restriction can attenuate inflammation and improve metabolic function (7, 21, 31, 35). Whether the beneficial impact of macrophage depletion extends beyond weight loss has not been examined.

Therefore, through the utilization of pair-feeding, our study was designed to determine if macrophage depletion could serve as an effective intervention to reduce obesity-driven adipose tissue inflammation and improve metabolic dysfunction beyond the weight loss seen with reduced energy intake.

METHODS

Animals.

Male wild-type C57BL/6J mice were purchased from the Jackson Laboratories (Bar Harbor, ME) and cared for at the Animal Resources Facility at the University of South Carolina. Mice (n = 15/group) were housed five per cage, maintained on a 12:12-h light-dark cycle in a low-stress environment (22°C, 50% humidity, low noise), and given water ad libitum. Principles of laboratory animal care were followed, and the Institutional Animal Care and Usage Committee of the University of South Carolina approved all experiments.

Diets.

At 6 wk of age, mice were randomly assigned to a control purified AIN-76A low-fat diet (LFD; 3.77 kcal/g) or a purified high-fat diet [HFD (40% of total kcal from fat); 4.57 kcal/g] designed to mimic the standard American diet (BioServ, Frenchtown, NJ). We used this diet in several of our previous investigations (11–16). Initially, the respective LFD or HFD was provided for 16 wk, at which time baseline body composition was determined (dual-energy X-ray absorptiometry) and metabolic tests were performed. Based on these outcomes, mice within each diet were separated into one of four treatment groups [LFD phosphate-buffered saline (PBS)-encapsulated liposomes, HFD PBS-encapsulated liposomes, LFD clodronate (CLD)-encapsulated liposomes, and HFD CLD-encapsulated liposomes] to match body weight and composition within each dietary treatment. The mice continued on their respective diets (LFD or HFD) for another 4 wk. Thus, in total, the mice consumed the diets for 20 wk.

CLD treatment and pair-feeding.

To deplete adipose tissue macrophages, a 200-μl intraperitoneal injection (~1 mg) of CLD-encapsulated liposomes (ClodronateLiposomes, Liposoma, Amsterdam, Netherlands) was administered twice a week beginning at week 16 of dietary treatment (8 injections in total) until euthanization at week 20. The selection of this dose of CLD was based on pilot work that we conducted. Intraperitoneal CLD-liposome administration was chosen, as this has been shown to be a more effective means of depleting adipose tissue macrophages than intravenous administration (27, 28). Initially, we used the manufacturer’s recommended dose of 0.05 mg/g body wt. However, the dose was scaled down to ≈0.03 mg/g body wt in an attempt to lessen toxicity but still elicit significant macrophage depletion, as we previously showed with respect to colonic macrophages (3). CLD treatment was initiated after 16 wk of HFD feeding, as we previously showed that 16 wk of HFD feeding induces an obese phenotype, characterized by adipose tissue inflammation with significant adipose tissue macrophage recruitment and metabolic dysfunction (12, 13). PBS-encapsulated liposomes (ClodronateLiposomes) were administered in the same manner as CLD-liposomes to serve as a control. The sample size in the LFD CLD and HFD CLD groups was reduced to 9 and 14 mice, respectively, due to apparent toxicity of CLD treatment. In our unpublished and published (3) experiments, we found that macrophage depletion reduces energy intake. This is in accordance with another published report that macrophage depletion (via the use of the diphtheria toxin) significantly reduced energy intake (29). Therefore, PBS-treated mice were pair-fed to their CLD-treated counterparts so as not to allow energy intake to be a potential confounding factor in the outcomes of the study. As the number of mice in each cage was equal across treatment groups, pair-feeding calculations were determined as follows: food consumption of CLD-treated animals was measured by averaging food intake per cage each day before food was administered to the counterpart PBS-treated groups. For example, if the three CLD HFD cages consumed 15, 16, and 16.4 g of food over a 24-h period, the average of this food consumption ( = 15.8 g) was provided to each of the three PBS HFD cages.

Body weight, energy intake, and body composition.

Body weight was monitored on a weekly basis throughout the study. Energy intake was monitored on a weekly basis before CLD treatment and on a daily basis upon commencement of CLD treatment for pair-feeding purposes. Energy intake was calculated as the total food consumption in grams per cage per week per mouse multiplied by the energy density per gram of the respective diet. Body composition was assessed before and after CLD treatments (weeks 16 and 20). For this procedure, mice were briefly anesthetized via isoflurane inhalation, and lean mass, fat mass, and percent body fat were assessed by dual-energy X-ray absorptiometry (Lunar PIXImus).

Metabolism.

Metabolic parameters were assessed before (16 wk of dietary treatment) and after (20 wk of dietary treatment/4 wk of CLD treatment) CLD treatment. After a 5-h fast, blood samples were collected from the tip of the tail. A glucometer (Bayer Contour, Mishawaka, IN) was used to determine blood glucose concentrations in whole blood. Collected blood was centrifuged at 4,000 rpm for 10 min at 4°C. Plasma was divided into aliquots and stored at −80°C until analysis. Plasma insulin concentrations were analyzed according to the manufacturer’s instructions using a mouse insulin ELISA kit (Mercodia, Winston Salem, NC). For glucose tolerance tests (GTTs) and insulin tolerance tests (ITTs), mice were fasted for 5 h, and glucose or insulin was administered intraperitoneally at 2 g/kg or 0.75 U/kg lean mass, respectively. A glucometer (Bayer Contour) was used to measure blood glucose concentrations (tail sampling) intermittently over a 2-h period (0, 15, 30, 60, 90, and 120 min) for GTTs and intermittently over a 1-h period (0, 15, 30, 45, and 60 min) for ITTs. Area under the curve (AUC) was calculated using the trapezoidal rule. Fasting serum was collected for free fatty acid (FFA) analysis at the 0- and 30-min time points of the post-CLD treatment ITT to assess insulin’s ability to inhibit lipolysis. FFAs were analyzed using a commercially available kit according to the manufacturer’s instructions (Wako Diagnostics, Richmond, VA).

Tissue collection.

After 20 wk of dietary treatment (4 wk after CLD treatment), mice were euthanized via isoflurane inhalation for tissue collection. An aliquot of whole blood taken from the inferior vena cava was analyzed for hematology using a VetScan Hm5 (Abaxis, Union City, CA). The remaining blood was spun at 4,000 rpm for 10 min, and plasma was stored at −80°C. Epididymal, mesenteric, and perirenal fat pads, as well as the liver, were removed, weighed, and immediately snap-frozen in liquid nitrogen and stored at −80°C or fixed in 4% formalin until analysis.

Hepatic lipid content and alanine transaminase activity assay.

Lipids were isolated from the liver utilizing a modified Folch extraction method and quantified gravimetrically, as previously described (13, 14, 16, 18, 44). Plasma alanine transaminase (ALT) activity was assessed using an Infinity ALT liquid reagent (Thermo Fisher, Middletown, VA) according to the manufacturer’s instructions.

Adipose tissue hydroxyproline assessment.

To assess the impact of macrophage depletion on adipose tissue fibrosis, epididymal adipose tissue hydroxyproline content [an indirect method for assessing collagen deposition (46)] was examined. Because of limitations in the amount of available epididymal adipose tissue, this assay was limited to the adipose tissue of HFD groups only. Briefly, epididymal adipose tissue (100–200 mg) was acid-hydrolyzed overnight at 110°C using 6 N HCl. The solution was centrifuged (10 min at 10,000 rpm), and the resulting supernatant was collected and dried under nitrogen gas. After incubation of the dried extract with 0.625 ml of a chloramine-T solution for 30 min, 0.625 ml of Ehrlich’s solution was added, as previously described (10). Samples were then incubated for 20 min at 65°C, cooled in room-temperature water, and transferred to a 96-well microplate and read using a spectrophotometer at 550 nm. Hydroxyproline (Sigma Aldrich, St. Louis, MO) was used to create a standard curve. Total hydroxyproline content was normalized to the epididymal adipose tissue weight.

Adipose tissue and liver histology.

Formalin-fixed tissues were embedded in paraffin blocks and sectioned. Epididymal and perirenal fat pads and the liver were stained with hematoxylin and eosin. Representative images were taken at ×20 and ×100 magnification. Antibodies, antibody dilutions, and antigen-retrieval methods for immunofluorescence and immunohistochemistry staining are provided in Table 1. For negative controls, a serial section was assayed exactly as described for the treated section, but without the primary antibody.

Table 1.

Primary and secondary antibodies as well as conditions used for immunofluorescence and immunohistochemistry

Primary Antibody (Vendor)	Dilution	Antigen Retrieval	Secondary Antibody/Incubation
Rat anti-mouse F4/80 (Bio-Rad, Hercules, CA)	1:200	Proteinase K (10 min)	Alexa Fluor 555 goat anti-rat  (Thermo Fisher, Waltham, MA)/1:200 at 37°C for 30 min
Rat anti-mouse Ly6G (BioLegend, San Diego, CA)	1:100	Heat-induced (citrate buffer, 10 mmol)	Goat anti-rat HRP-conjugated  (Abcam, Cambridge, UK)/1:200 at RT for 1 h*
Rat anti-mouse F4/80 (Bio-Rad, Hercules, CA)	1:200	Proteinase K (10 min)	Goat anti-rat HRP-conjugated  (Abcam, Cambridge, UK)/1:200 at RT for 1 h*

All primary antibodies were incubated overnight at 4°C. HRP, horseradish peroxidase; RT, room temperature.

^*3,3′-Diaminobenzidine was used for visualization.

Quantitative real-time RT-PCR.

An RNeasy Lipid Tissue Mini Kit (Qiagen, Valencia, CA) was used to isolate RNA from epididymal and perirenal adipose tissue depots, as well as the liver. TaqMan reverse transcription reagents and gene expression assays (Applied Biosystems, Foster City, CA) were used to reverse transcribe and analyze expression of the following genes: F4/80, CD68, CD11c, CD206, MCP-1, TNFα, IL-10, IL-6, and IL-1β. Potential reference genes [ribosomal protein lateral stalk subunit P2 (RPLP2), 18s rRNA (18s), hydroxymethylbilane synthase (HMBS), TATA-box binding protein (TBP), β₂-microglobulin (B2M), H2A histone family member V (H2AFV), and hypoxanthine phosphoribosyltransferase (HPRT)] were analyzed for stability using Qbase+ software (Biogazelle, Ghent, Belgium) for each tissue analyzed. The optimal number of reference genes was determined by Qbase+, and the geometric mean of these genes was used as the normalization factor for each analysis: epididymal adipose tissue (HMBS, TBP, B2M, and H2AFV), perirenal adipose tissue (HMBS and HPRT), and the liver (TBS and B2M) (43). Gene expression was quantified using the ΔΔC_T method and Qbase+ software.

Western blot analysis.

Briefly, epididymal and perirenal adipose tissues, as well as a portion of the liver, were homogenized in Mueller buffer containing a protease inhibitor cocktail (Sigma Aldrich) (8). Total protein concentrations were determined by the Bradford method. Equal amounts of crude protein homogenates were fractioned on hand-casted 12% SDS-polyacrylamide gel and electrophoretically transferred to a polyvinylidene difluoride membrane using a Genie Blotter (IDEA Scientific, Minneapolis, MN). Membranes were stained with a Ponceau S solution to verify equal protein loading and transfer efficiency. Subsequently, membranes were blocked for 1 h in 5% milk in Tris-buffered saline-0.1% Tween 20 (TBST). The lymphocyte antigen 6G (Ly6G) primary antibody (Abcam, Cambridge, UK) was diluted 1:1,000 in 5% milk-TBST overnight at 4°C. An anti-rat (Cell Signaling Technology, Danvers, MA) IgG horseradish peroxidase-conjugated secondary antibody was diluted 1:2,000 in 5% milk-TBST and incubated for 1 h at room temperature. An enhanced chemiluminescence substrate for detection of horseradish peroxidase (Thermo Scientific, Watham, MA) was used to visualize the antibody-antigen interaction. Autoradiography films were scanned and blots were quantified using scientific imaging software (ImageJ). Because of tissue limitations, only the perirenal fat from HFD-fed animals was analyzed for Ly6G content. After Western blotting, all membranes were stained with amido black (3, 13, 44), and the densitometry of each lane was calculated using Quantity One software (Bio-Rad, Hercules, CA), allowing for total protein normalization. This method of normalization has been shown to be more accurate than typically used loading controls (2).

Plasma cytokines.

Plasma collected at euthanization was analyzed for circulating IL-6 and MCP-1 (LEGEND MAX ELISA kits, BioLegend, San Diego, CA). Because of limitations in the amount of plasma, only the plasma of the HFD-fed groups was analyzed for MCP-1 concentration.

Statistical analysis.

Data were analyzed using commercially available statistical software: Prism 6 (GraphPad Software, La Jolla, CA) and SigmaStat (Systat Software, San Jose, CA). A two-way ANOVA followed by a Newman-Keuls post hoc analysis was used to determine differences between diet (HFD vs. LFD) and treatment (CLD vs. PBS). A two-way (time × group) repeated-measures ANOVA followed by a Newman-Keuls post hoc test was used for body weight analysis. A two-way (time × group) repeated-measures ANOVA followed by a Newman-Keuls post hoc test was also used for ITT analysis to determine differences in blood glucose levels from baseline (0 min) over the duration of the test. A Student’s t-test (2-tailed) was used for Western blot analysis of Ly6G, epididymal adipose tissue hydroxyproline content, and plasma MCP-1 concentrations to compare treatment (CLD vs. PBS) within diet. Given the small sample size (n = 3/group), statistical analysis of energy intake was not assessed. Any statistical test that did not pass the equal-variance test (Bartlett’s test for equal variances) was log-transformed and then reanalyzed. Data are presented as means ± SE, and the level of significance was set at P < 0.05.

RESULTS

HFD consumption for 16 wk leads to an obese phenotype.

Prior to CLD or PBS treatment, mice were fed the 40% HFD or the LFD for 16 wk. An obese phenotype was successfully achieved, as the HFD-fed mice gained significantly more body weight and fat mass and displayed a significantly greater percentage of body fat (Fig. 1, A and B) than LFD-fed mice (P < 0.05). HFD-fed mice also displayed elevated fasting blood glucose, insulin, and GTT AUC and impaired insulin sensitivity (Fig. 1, C–F), demonstrating a dysfunctional metabolic phenotype (P < 0.05).

Fig. 1.

Body composition and metabolic data after 16 wk of dietary treatment [high-fat diet (HFD) or low-fat diet (LFD)] before start of liposome treatment. Mice were fed the HFD or LFD for 16 wk. A: body weight. B: body composition. C: fasting (5-h) blood glucose levels. D: fasting (5-h) plasma insulin levels. E: intraperitoneal glucose tolerance test (GTT) and corresponding area under the curve (AUC). F: intraperitoneal insulin tolerance test (ITT) following a 5-h fast. CLD, clodronate. Values are means ± SE; n = 9–15 mice per group. *P < 0.05, HFD vs. LFD groups. Bar graphs not sharing a common letter (a, b) are significantly different from one another (P < 0.05). %, $, #, ^P < 0.05 vs. baseline (0 min) for LFD PBS, LFD CLD, HFD PBS, and HFD CLD groups, respectively.

CLD treatment reduces body weight by eliciting changes in energy intake.

After 16 wk on the HFD or LFD, PBS or CLD-encapsulated liposomes were administered for a 4-wk period. The PBS-treated groups were pair-fed the same amount of food as their CLD-treated counterparts. All CLD-treated groups displayed a decrease in body weight starting 1 wk after CLD treatment (Fig. 2A) that was accompanied by a decrease in energy intake (Fig. 2B), suggesting that CLD treatment induced a reduction in appetite. There was a main effect of the HFD to increase lean mass, fat mass, and percent body fat relative to the LFD-fed mice (Fig. 2C; P < 0.05). Pair-feeding was successful in matching the body weight, fat mass, and lean mass in the PBS and CLD treatment groups. However, within the LFD groups, CLD treatment significantly reduced percent body fat (Fig. 2C; P < 0.05). The HFD increased all fat pad weights (epididymal, mesenteric, and perirenal) relative to the LFD-fed mice (Fig. 2D; P < 0.05). Interestingly, within the HFD-fed mice, CLD-treated mice exhibited an increase in epididymal, but not perirenal or mesenteric, fat pad weight. As a consequence, the sum of all three fat pads was also increased (Fig. 2D; P < 0.05). We found this surprising, and to potentially explain this discrepancy, we examined a marker of extracellular matrix (collagen) accumulation (hydroxyproline content) in epididymal adipose tissue (HFD groups only). However, no significant difference was found in hydroxyproline content (data not shown) between these groups, suggesting that some other mechanisms are responsible for the epididymal fat pad weight between the HFD groups.

Fig. 2.

Body morphology following dietary [high-fat diet (HFD) or low-fat diet (LFD)] and liposome treatment. A: body weight. B: energy intake throughout the experimental protocol. C: terminal body composition analysis as assessed by dual-energy X-ray absorptiometry. D: intraperitoneal adipose tissue depot weights (epididymal, perirenal, mesenteric), and total intraperitoneal fat weight measured at euthanization. CLD, clodronate. Values are means ± SE; n = 9–15 mice per group. *Significantly different from start of liposome treatment at 16 wk. Bar graphs not sharing a common letter (a, b, c) are significantly different from one another (P < 0.05).

CLD-liposome treatment does not reduce epididymal adipose tissue macrophage populations in a HFD setting, yet it results in an exacerbated cytokine profile and significant increases in circulating and infiltrating neutrophils.

We initially examined epididymal adipose tissue to assess the effectiveness of CLD treatment to deplete macrophages, as the epididymal fat pad has been shown to develop signs of inflammation and insulin resistance before other intraperitoneal fat pads (39). Interestingly, while, as expected, the HFD increased adipose tissue macrophage markers, no significant difference was found with respect to gene expression of overall macrophage markers (F4/80 and CD68) or M1 (CD11c) and M2 (CD206) macrophage markers between the HFD groups (Fig. 3A). This lack of difference in macrophage populations was substantiated by immunofluorescence staining in the epididymal fat pad (Fig. 3B). However, within the LFD groups, CLD treatment significantly decreased F4/80 (−58%), CD68 (−52%), and CD206 (−72%) gene expression, but not CD11c mRNA expression (Fig. 3A; P < 0.05).

Fig. 3.

Treatment with clodronate (CLD) for 4 wk depletes epididymal adipose tissue macrophages in a low-fat diet (LFD), but not a high-fat diet (HFD), setting. A: epididymal adipose tissue mRNA expression of macrophage markers (F4/80, CD68, CD11c, and CD206) normalized to the most stable internal reference genes [hydroxymethylbilane synthase (HMBS), H2A histone family member V (H2AFV), β₂-microglobulin (B2M), and TATA-box binding protein (TBP)] calculated using Qbase+ software. B: representative immunofluorescence images (×40) of DAPI and F4/80 in epididymal adipose tissue of HFD-fed mice. Values are means ± SE; n = 9–15 mice per group. Bar graphs not sharing a common letter (a, b, c) are significantly different from one another (P < 0.05).

We then proceeded to examine the gene expression of various inflammatory cytokines in the epididymal fat pad (Fig. 4A; P < 0.05). As expected, the HFD significantly increased the gene expression of the monocyte chemoattractant MCP-1, the proinflammatory cytokine TNF-α, and the anti-inflammatory cytokine IL-10 relative to the LFD groups (P < 0.05). However, within the HFD groups, CLD treatment further elevated gene expression of MCP-1 and IL-10 (P < 0.05). Of additional interest was the finding that CLD treatment, independent of diet, resulted in substantial increases in mRNA expression of the proinflammatory cytokines IL-6 and IL-1β (P < 0.05). In fact, IL-1β mRNA induction was further augmented in the CLD LFD group relative to the CLD HFD group, whereas IL-1β mRNA expression was downregulated in the HFD PBS group relative to the LFD PBS group (P < 0.05).

Fig. 4.

Despite not affecting epididymal adipose tissue macrophage populations in a high-fat diet (HFD) setting, clodronate (CLD) treatment augments epididymal adipose tissue inflammation and polymorphonuclear cell infiltration. A: epididymal adipose tissue mRNA expression of inflammatory mediators [monocyte chemoattractant protein-1 (MCP-1), TNF-α, IL-10, IL-6, and IL-1β] normalized to the geometric mean of the most stable internal reference genes [hydroxymethylbilane synthase (HMBS), H2A histone family member V (H2AFV), β₂-microglobulin (B2M), and TATA-box binding protein (TBP)] calculated using Qbase+ software. B: representative images (×20 and ×100) of hematoxylin-eosin-stained epididymal adipose tissue sections showing increased polymorphonuclear cell infiltration in the CLD-treated groups. LFD, low-fat diet. Values are means ± SE; n = 9–15 mice per group. Bar graphs not sharing a common letter (a, b, c) are significantly different from one another (P < 0.05).

Based on these findings, we performed hematoxylin-and-eosin staining to histologically assess the epididymal fat pads (Fig. 4B). We found clear evidence of infiltrating polymorphonuclear cells (PMCs) in the CLD-treated groups relative to the PBS-treated groups. Additionally, in whole blood collected at euthanization, we found significant increases in circulating neutrophils in the CLD-treated groups (Fig. 5A). This finding led us to believe that the tissue PMCs we examined histologically were neutrophils. Using Western blotting for Ly6G, a neutrophil marker, we found significant increases in Ly6G in the CLD-treated groups relative to their PBS controls (Fig. 5B; P < 0.05). This discovery was corroborated by immunohistochemistry staining of Ly6G (Fig. 5C).

Fig. 5.

Clodronate (CLD) treatment increases circulating and epididymal adipose tissue neutrophils. A: circulating neutrophils determined in whole blood by a VetScan HM5. B: Western blot analysis of lymphocyte antigen 6G (Ly6G, a neutrophil marker) in epididymal adipose tissue normalized to total protein stain (amido black). IOD, integrated optical density. C: 2 representative images of immunohistochemistry staining (×100) of Ly6G in epididymal adipose tissue. HFD, high-fat diet; LFD, low-fat diet. Values are means ± SE; n = 9–15 mice per group. *P < 0.05. Bar graphs not sharing a common letter (a, b) are significantly different from one another (P < 0.05).

Macrophage populations are significantly diminished in perirenal adipose tissue, yet a proinflammatory environment, including neutrophil infiltration, is evident.

Because we did not find significant decreases in macrophage populations in the epididymal fat pad, we decided to assess the perirenal fat pad with respect to macrophage infiltration. Contrary to the finding in the epididymal fat pad, CLD treatment significantly decreased F4/80 (−30%), CD68 (−35%), CD11c (−40%), and CD206 (−42%) in the CLD HFD group relative to the PBS HFD group (Fig. 6A; P < 0.05), which was corroborated by immunofluorescence staining for F4/80 (Fig. 6B). In the LFD groups, a similar pattern was found, as CLD treatment reduced mRNA content of F4/80 (−43%), CD68 (−35%), and CD206 (−47%) (Fig. 6A; P < 0.05). However, similar to the epididymal adipose tissue, CLD treatment did not influence CD11c gene expression.

Fig. 6.

Treatment with clodronate (CLD) for 4 wk depletes macrophages in the perirenal fat pad. A: perirenal adipose tissue mRNA expression of macrophage markers (F4/80, CD68, CD11c, and CD206) normalized to the geometric mean of the most stable internal reference genes [hydroxymethylbilane synthase (HMBS) and hypoxanthine phosphoribosyltransferase (HPRT)] using QBase+ software. HFD, high-fat diet; LFD, low-fat diet. B: representative immunofluorescence images (×40) of DAPI and F4/80 in perirenal adipose tissue of HFD-fed mice. Values are means ± SE; n = 9–15 mice per group. Bar graphs not sharing a common letter are significantly different from one another (P < 0.05).

Since we found reduced macrophage populations in the CLD-treated groups, we moved on to explore the gene expression of several proinflammatory cytokines (TNF-α, MCP-1, IL-1β, and IL-6). As expected, we found a HFD effect to increase TNF-α and MCP-1 expression (Fig. 7A; P < 0.05). On the other hand, IL-6, similar to the finding in epididymal adipose tissue, was downregulated as a result of HFD consumption (P < 0.05). In the HFD setting, CLD treatment significantly decreased TNF-α gene expression (−31%), which corresponded with the similar decrease (−30%) in F4/80 mRNA (Fig. 7A; P < 0.05). However, analogous to our observation in the epididymal adipose tissue, MCP-1, IL-1β, and IL-6 were increased as a result of CLD treatment among the HFD groups (Fig. 7A; P < 0.05). A similar finding was observed with CLD treatment in the LFD setting, as IL-1β and IL-6 expression were significantly increased (P < 0.05). This outcome, paired with no change in MCP-1 gene expression, paralleled observations in epididymal adipose tissue.

Fig. 7.

An increase in perirenal adipose tissue inflammation is paired with an increase in infiltrating neutrophils. A: perirenal adipose tissue mRNA expression of inflammatory mediators [TNF-α, monocyte chemoattractant protein-1 (MCP-1), IL-1β, and IL-6] normalized to the geometric mean of the most stable internal reference genes [hydroxymethylbilane synthase (HMBS) and hypoxanthine phosphoribosyltransferase (HPRT)] using QBase+ software. HFD, high-fat diet; LFD, low-fat diet. B: 2 representative images of immunohistochemistry staining (×100) of lymphocyte antigen 6G (Ly6G) in perirenal adipose tissue of HFD-fed groups. C: Western blot analysis of Ly6G in epididymal adipose tissue normalized to total protein stain (amido black). IOD, integrated optical density. Values are means ± SE; n = 9–15 mice per group. *P < 0.05. Bar graphs not sharing a common letter (a, b, c, d) are significantly different from one another (P < 0.05).

We followed up the quantitative RT-PCR analysis with a histological assessment of the perirenal fat pad and, similar to our observation in the epididymal fat pad, discovered an abundance of infiltrating PMCs surrounding adipocytes. Immunohistochemistry confirmed that these cells were Ly6G-positive (Fig. 7B). Western blot analysis of Ly6G verified that Ly6G content was greater in the CLD-treated mice (Fig. 7C; P < 0.05).

CLD treatment does not regress early-stage NAFLD development.

HFD treatment resulted in early-stage NAFLD development, as characterized by hepatomegaly (Fig. 8A), hepatic steatosis (Fig. 8, B and D), and elevated plasma ALT (Fig. 8C; P < 0.05). However, the disease did not progress to steatohepatitis as in the HFD PBS group, although it did display an increase in F4/80 expression relative to the LFD PBS group (Fig. 8, E and F; P < 0.05) but did not histologically present significant evidence of immune cell infiltration. Moreover, there were no significant increases in any of the proinflammatory cytokines, including MCP-1, TNF-α, IL-6, and IL-1β (Fig. 8G). In fact, there was a main effect for HFD consumption to decrease MCP-1 gene expression, and the HFD PBS group displayed less TNF-α gene expression than the LFD PBS group (P < 0.05).

Fig. 8.

Clodronate (CLD) treatment does not affect regression of early-stage nonalcoholic fatty liver disease development. A: liver weight following euthanization. HFD, high-fat diet; LFD, low-fat diet. B: hepatic lipid accumulation. C: plasma alanine transaminase (ALT) activity. D: representative hepatic hematoxylin-eosin-stained (×20) images. E: hepatic gene expression of F4/80. F: representative F4/80 staining (×60, arrows indicate examples of positive staining). G: mRNA expression of hepatic inflammatory mediators [monocyte chemoattractant protein-1 (MCP-1), TNF-α, IL-6, and IL-1β] normalized to the geometric mean of the most stable internal reference genes [β₂-microglobulin (B2M) and TATA-box binding protein (TBP)] using QBase+ software. ME, main effect. Values are means ± SE; n = 9–15 mice per group. Bar graphs not sharing a common letter (a, b, c) are significantly different from one another (P < 0.05).

Regardless of the lack of HFD-induced hepatic inflammation, CLD treatment significantly reduced gene expression of F4/80 (−59%) in a HFD setting (Fig. 8E). This was substantiated by F4/80 staining (Fig. 8F) and occurred independent of any changes in liver weight or hepatic lipid accumulation. Surprisingly, this decrease in F4/80 did not impact gene expression of MCP-1 or TNF-α (Fig. 8G). Additionally, CLD treatment in a LFD setting did not significantly reduce macrophages, as substantiated by F4/80 gene expression and immunohistochemistry (Figs. 8, E and F), yet there was a main effect for CLD treatment to decrease TNF-α and IL-1β (Fig. 8G; P < 0.05). Despite these reductions in TNF-α and IL-1β, CLD treatment, independent of diet, resulted in higher IL-6 gene expression (Fig. 8G; P < 0.05).

Macrophage depletion elicits significant increases in circulating IL-6 and MCP-1.

Because we found a main effect of CLD treatment to increase gene expression of IL-6 in all tissues analyzed and an interaction for CLD treatment to increase epididymal and perirenal MCP-1 gene expression among the HFD groups, we examined circulating IL-6 and MCP-1 levels (Fig. 9). Consistent with the gene expression, CLD treatment, irrespective of diet, induced significant increases in plasma IL-6 concentration (Fig. 9A; P < 0.05). Additionally, CLD-treated HFD mice displayed a more than twofold increase in circulating MCP-1 relative to HFD PBS mice (Fig. 9B; P < 0.05).

Fig. 9.

Clodronate (CLD) treatment significantly increases circulating proinflammatory cytokine concentrations. After 20 wk of diet/4 wk of CLD-liposome treatment, plasma was assessed for circulating proinflammatory cytokines. A: IL-6. B: monocyte chemoattractant protein-1 (MCP-1). Because of limitations in the amount of plasma, only the plasma of high-fat-diet (HFD)-treated groups was assessed for concentration of circulating MCP-1. LFD, low-fat diet. Values are means ± SE; n = 7–14 mice per group. Bar graphs not sharing a common letter (a, b) are significantly different from one another (P < 0.05).

Macrophage depletion does not rescue impaired glucose metabolism or insulin resistance and exacerbates adipose tissue insulin action.

CLD treatment did not improve glucose metabolism or insulin resistance, as fasting blood glucose (Fig. 10A), fasting insulin (Fig. 10B), and GTT AUC (Fig. 10C) were not reduced with CLD treatment in the LFD and HFD groups. Similarly, insulin’s ability to decrease blood glucose levels (Fig. 10D) was not enhanced with CLD treatment in the LFD or HFD setting. With respect to FFA metabolism, basal serum FFAs were significantly increased in the LFD PBS group compared with all other groups (Fig. 10E; P < 0.05). At 30 min after exogenous insulin administration, there was a main effect of diet (P < 0.05) for the LFD-fed mice to exhibit significantly lower serum FFAs than HFD-fed mice (Fig. 10F). When the absolute change in serum FFAs over the 30-min period was examined, there was a main effect of CLD treatment to impair the decrease in serum FFAs compared with PBS treatment. Among the groups, the LFD PBS group displayed the greatest drop in serum FFAs (Fig. 10G; P < 0.05). LFD CLD mice exhibited a more significant drop in serum FFAs than the HFD CLD mice (P < 0.05), but not when compared with the HFD PBS mice. Although not statistically significant, there was a trend (P = 0.08) for a greater decrease in serum FFAs in the HFD PBS than the HFD CLD group.

Fig. 10.

Clodronate (CLD) treatment does not rescue impaired glucose metabolism or insulin resistance and exacerbates adipose tissue insulin action. After 20 wk of diet/4 wk of CLD-liposome treatment, metabolic outcomes were assessed. A: fasting (5-h) blood glucose levels. B: fasting (5-h) plasma insulin levels. C: intraperitoneal glucose tolerance test (GTT) and corresponding area under the curve (AUC). D: intraperitoneal insulin tolerance test (ITT) following a 5-h fast. E and F: serum free fatty acid concentration following a 5-h fast measured at 0 and 30 min of the ITT. G: change in serum free fatty acids from 0 to 30 min of the ITT. HFD, high-fat diet; LFD, low-fat diet; ME, mixed effects. Values are means ± SE; n = 9–15 mice per group. Bar graphs not sharing a common letter (a, b, c) are significantly different from one another (P < 0.05). %, $, #, ^P < 0.05 vs. baseline (0 min) for LFD PBS, LFD CLD, HFD PBS, and HFD CLD groups, respectively.

CLD treatment results in anemia.

It is well established that macrophages play a central role in heme-iron metabolism (19, 33, 40). Therefore, we analyzed whole blood to see if CLD treatment impacted hemoglobin levels. Data from the hematology analysis showed that CLD treatment, independent of diet, decreased absolute hemoglobin levels and the hematocrit (Table 2; P < 0.05).

Table 2.

CLD treatment results in anemia

	Treatment
	LFD PBS	LFD CLD	HFD PBS	HFD CLD
Mean corpuscular hemoglobin concentration, g/l	36.5 ± 0.42	35.1 ± 0.35	35.2 ± 0.23	35.6 ± 0.59
Hemoglobin, g/dl	13.7 ± 0.21a	11.6 ± 0.22b	13.2 ± 0.18a	11.8 ± 0.24b
Hematocrit, %	37.7 ± 0.62a	33 ± 0.55b	37.5 ± 0.52a	33.3 ± 0.81b

Values are means ± SE; n = 9–15 mice per group. Whole blood collected at euthanization (4 wk after liposome treatment) was assessed for mean corpuscular hemoglobin concentration, as well as hemoglobin and %hematocrit. CLD, clodronate; HFD, high-fat diet; LFD, low-fat diet.

^a,bGroups not sharing a common letter differ significantly from one another (P < 0.05).

DISCUSSION

Because of their prominent role in obesity-associated inflammation, macrophages have become the target for therapies aimed at mitigating the chronic inflammatory environment and impaired metabolic processes. As others have reported, adipose tissue (perirenal) macrophages were successfully reduced with CLD-liposome treatment (6, 17, 27, 30). However, relative to weight-matched control animals, CLD treatment did not rescue HFD-induced impaired glucose metabolism. Moreover, CLD treatment, irrespective of diet, increased circulating and adipose tissue infiltrating neutrophils. Additionally, we found that CLD treatment resulted in reduced energy intake and significant body weight loss.

Interestingly, others who utilized CLD-liposomes to decrease obesity-induced adipose tissue macrophage infiltration have not reported similar findings of neutrophilia, an inability to rescue impaired metabolism, and reduced energy intake (6, 17, 25, 30). However, our results are consistent with those of Lee et al., who found that macrophage depletion utilizing a lysozyme M promoter-directed Cre (LysM^Cre) diphtheria toxin model resulted in significant neutrophil infiltration, decreased energy intake, and reduced body weight in both HFD- and standard chow-fed mice (29). Similarly, Gordy et al. developed a LysM^Cre floxed FLICE-like inhibitory protein (LysM^Cre c-FLIP^f/f) mouse model, which results in mice lacking splenic marginal zone, bone marrow, and thioglycollate-elicited peritoneal macrophages, and found that neutrophilia is secondary to macrophage depletion and that macrophages regulate neutrophil homeostasis (20). Although energy intake was not measured in the study by Gordy et al., it was discovered that these mice exhibit a decreased body weight. Similar results of neutrophilia and a loss of body weight were reported by Wu et al. when conditional macrophage depletion via the use of a transgenic mouse model was used to study the impact of macrophages on osteoarthritis in obesity (47). The results of Lee et al., Gordy et al., and Wu et al. and the mouse models utilized in their studies provide support that our findings are a direct result of macrophage depletion and not of CLD-liposome treatment alone.

It was surprising that we did not find significant decreases in macrophage populations in the epididymal adipose tissue of the HFD CLD group. It was our initial hypothesis that epididymal adipose tissue would be the most sensitive to CLD treatment, given the route of liposome administration and exacerbated inflammation and impaired insulin action at an earlier time point in the epididymal than the other intraperitoneal fat pads (39). However, the fact that CLD treatment significantly depleted epididymal and perirenal adipose tissue macrophages in a LFD setting and resulted in significant perirenal adipose tissue macrophage depletion in a HFD setting leads us to believe that perhaps epididymal adipose macrophages are more sensitive to the pair-feeding. Prior work demonstrated that restriction of food intake in obese animals reverses macrophage accumulation in perigonadal adipose tissue (24). Inasmuch as we are comparing the CLD treatment with pair-fed animals that lost weight, it is possible that CLD could not reduce macrophage populations beyond that seen with caloric restriction. Alternatively, the inability to detect a difference in epididymal adipose tissue macrophage populations may be due to the immune system’s ability to compensate for the effects of CLD treatment to recruit more monocytes to the inflamed area, resulting in a repopulation of macrophages. This idea is supported by the significant increase in epididymal mRNA expression of the potent monocyte chemokine MCP-1, as previously shown with acute intraperitoneal CLD-liposome treatment (28), a higher concentration of circulating MCP-1, and tissue neutrophilia in the HFD CLD group. However, a time-course study with CLD treatment would be necessary to confirm this.

It is important to consider that dendritic cells, another phagocytic immune cell, are also known to be affected by CLD-liposome treatment (42). Thus the neutrophil infiltration makes physiological sense, as they are the only primary professional phagocytic cell remaining after macrophage and dendritic cell depletion. Neutrophil infiltration into the adipose tissue serves two probable functions: 1) removal of dead immune cells affected by CLD treatment and 2) adoption of the role of macrophages by cleaning up dead adipocytes. This is supported by the observation that Ly6G-positive cells surround adipocytes in the CLD-treated mice. Additionally, it is likely that the increased expression of adipose tissue IL-6 and IL-1β, as well as the circulating IL-6 concentration, originates from these infiltrating neutrophils, as the HFD PBS treatment did not induce significant increases in these cytokines relative to the LFD PBS group. Furthermore, Lee et al. and Gordy et al. found that macrophage depletion-induced neutrophilia resulted in significant increases in circulating plasma cytokines, including IL-6 (20, 29) and IL-1β (20). However, we cannot rule out the possibility that other immune cells, such as T cells, or adipocytes could be producing these cytokines.

There is a clear discrepancy between the results of our study and the studies of others who have shown that macrophage depletion improves metabolic processes through the use of CLD treatment (6, 9, 17, 25, 30). In particular, our results are in direct contradiction to those of Lee et. al., who suggested that “long-term” obesity-associated insulin resistance, as would be induced after 16 wk on a HFD, is largely mediated by macrophage-induced proinflammatory actions (30). These discrepancies may be explained by many factors, including, but not limited to, the type of HFD utilized, the mode of CLD administration (intraperitoneal vs. intravenous), and the dose, timing, and duration of CLD treatment. For instance, the mode of CLD-liposome administration can have a significant effect on the tissue specificity of macrophage depletion. Lanthier et al. showed that intravenous CLD-liposome administration selectively depletes Kupffer cells (liver macrophages), whereas intraperitoneal administration is more advantageous for adipose tissue macrophage depletion (26, 28). The majority of previous studies employing CLD-liposomes to deplete adipose tissue macrophages have utilized intraperitoneal administration (6, 9, 17, 25, 30), as was used in our experiment. However, the CLD dose (≈0.03 mg/g body wt) utilized in our study was significantly less than that used in other studies (≈0.1 mg/g body wt) (6, 17, 30). Furthermore, the degree to which the duration of CLD-liposome treatment impacts the effectiveness of macrophage depletion and the resulting effect on metabolic function are conflicting. Studies that have employed intraperitoneal CLD-liposome administration for a relatively short period (1–2 wk), such as those of Feng et al. (17), Choe et al. (9) (2 treatments/wk for 1 wk), and Lee et. al. [“5 injections every 3 days for 14 days” (as described in supplemental data for their study)] (30), showed effective macrophage depletion and beneficial effects on metabolic outcomes. The duration (2 treatments/wk for 4 wk) of CLD-liposome administration was significantly longer in our study than in these previous studies and may be largely responsible for the discrepancy in outcomes between studies. However, the duration CLD-liposome treatment was significantly longer (2 treatments/wk for 9 wk) in the study of Bu et al. than in our study, yet they showed beneficial effects on adiposity and metabolism (6).

A potential confounding factor for the majority of the previous studies utilizing intraperitoneal CLD-liposome treatment, however, is the failure to examine/report energy intake or body weight changes with CLD treatment (9, 17, 30). Moreover, those studies that did report significant changes in body weight did not report energy intake, as exemplified in a recent publication by Kumar et al., who used a CLD-liposome dose similar to that utilized in our study (≈0.03 mg/g body wt administered weekly for 4 wk) and found that CLD-liposome treatment significantly depleted adipose tissue macrophages, improved metabolic outcomes, and dramatically reduced adiposity by >50% (25). Bu et al. reported food intake; however, they did not specify the number of mice housed per cage, the means by which energy intake was calculated if multiple mice were housed per cage, and the statistical analysis utilized to assess a potential difference in energy intake (6). Thus, in these previous studies, any potential benefits with respect to decreases in adiposity, inflammation, insulin resistance, and hepatic steatosis as a consequence of macrophage depletion may simply be a secondary effect resulting from decreased energy intake.

Although it was beyond the scope of our investigation, it is likely that the reduced energy intake and, consequently, body weight loss resulting from CLD-liposome treatment are due to neutrophilia, an exacerbated circulating proinflammatory environment, and brain inflammation. However, it cannot be ruled out that liposome treatment alone may have reduced food intake, as previously discussed (26). In our previous experience, we have not found an effect of PBS-liposome treatment on body weight or food intake. Moreover, the concept that reduced energy intake is due to CLD-liposome treatment and subsequent macrophage depletion, and not liposome treatment, in general, is supported by the findings of Lee et al., who reported hypothalamic inflammation and reduced energy intake in their macrophage-depleted mice (29). It has also been shown that IL-1β plays a key role in regulating this effect, as Gordy et al. found that administration of IL-1 receptor antagonist to macrophage-deficient mice rescued the decreased body weight (20). Consistent with this, in the present investigation and in that of Lee et al., decreased energy intake was observed with macrophage depletion, as were increases in adipose tissue IL-1β gene expression, which Lee et al. also found to be elevated in the hypothalamus (29).

Others have shown that hepatic resident macrophage (Kupffer cell) activation impairs insulin signaling in the early stages of HFD consumption (28). We assessed whether CLD treatment reduced Kupffer cell accumulation in our model. Although there was evidence of NAFLD, as assessed by hepatic steatosis, elevated plasma ALT activity, and increased macrophage content, the HFD did not induce robust hepatitis, as substantiated by no significant increases in various inflammatory markers (MCP-1, TNF-α, IL-6, and IL-1β). It is likely that a longer feeding period would be necessary to develop a more advanced NAFLD phenotype. Nonetheless, CLD treatment significantly decreased macrophages (F4/80) in a HFD setting. However, glucose metabolism was not improved. Lanthier et al. found similar results and concluded that Kupffer cell depletion does not have a therapeutic effect on detrimental metabolic changes induced by a HFD (27). Thus our results provide further support that ectopic lipid accumulation resulting from a positive energy balance may be a more important instigator of impaired metabolic processes than inflammation, as previously shown (41).

A limitation of our study is that CLD treatment is a nonspecific means of depleting macrophage populations. The existence of different macrophage subsets, generally categorized into two extreme phenotypes, pro- and anti-inflammatory, has been established. Although this dichotomy does not fully reflect the heterogeneity of macrophage populations, it is a widely used distinguishing characteristic. Therapies specifically targeting individual macrophage phenotypes or a specific tissue macrophage may be more effective than general systemic macrophage depletion (36). However, the establishment of such models and tissue specificity is still in its infancy.

It is important to note that CLD-treated groups displayed symptoms of anemia. Macrophages play an essential role in heme and iron metabolism (19, 33, 40). Iron can also modulate glucose homeostasis and adipose tissue function (23). Thus it is imperative that a complete blood count be part of a routine assessment when interventions aimed at modulating macrophage populations are utilized. This is especially true in the case of obesity, which is emerging as a risk factor for iron deficiency (1). It is also evident from our results that severe macrophage depletion and subsequent compensatory immune responses can be lethal in certain conditions. This is consistent with other macrophage-depletion models (32, 37). We hypothesize that the greater degree of toxicity in the LFD CLD group is the result of an immunocompromised state from severe macrophage depletion not reached in the HFD CLD group. This is likely the case, given that adipose tissue macrophage populations would have been significantly lower in the LFD- than HFD-fed mice after 16 wk of dietary treatment, as we previously showed (12, 13). Future assessments to evaluate potential toxicity of macrophage depletion, even in the short term, are necessary before any clinical interventions aimed at macrophage diminution.

In conclusion, our study shows that, when controlling for energy intake, macrophage depletion using CLD in an obese setting is not an effective therapy for rescuing metabolic dysfunction and may increase the risk for anemia, neutrophilia, and exacerbated adipose tissue inflammation. Thus the utilization of prolonged repeated intraperitoneal CLD-liposome treatment to deplete macrophages does not seem to be a viable option as a potential clinical therapy. Future studies are necessary to better understand the role of the timing, tissue specificity, and phenotype of macrophage depletion in obesity outcomes.

GRANTS

This work was funded by NIH Grants 1R21 CA-191966 (E. A. Murphy and C. M. Robinson) and 1F99 CA-234920-01 (J. E. Bader) and American Institute of Cancer Research Grant 359566 (E. A. Murphy).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.E.B., R.T.E., and E.A.M. conceived and designed research; J.E.B., R.T.E., and M.S.C. performed experiments; J.E.B., R.T.E., M.S.C., and E.A.M. analyzed data; J.E.B., R.T.E., K.T.V., M.S.C., O.P.M., and E.A.M. interpreted results of experiments; J.E.B., R.T.E., and M.S.C. prepared figures; J.E.B. and R.T.E. drafted manuscript; J.E.B., R.T.E., K.T.V., M.S.C., A.T.S., O.P.M., C.M.R., and E.A.M. edited and revised manuscript; J.E.B., R.T.E., K.T.V., M.S.C., A.T.S., O.P.M., C.M.R., and E.A.M. approved final version of manuscript.