High-fat diet or low-protein diet changes peritoneal macrophages function in mice

Obesity and protein malnutrition are major food problems nowadays, affecting billions of people around the world. The nutrition transition that has occurred in recent decades is changing the nutritional profile, reducing malnutrition and increasing the percentage of obese people. The innate immune response is greatly influenced by diet, with significant changes in both malnutrition and obesity. Therefore, we investigate the effects of protein malnutrition and obesity in nutritional and immunological parameters in mice. Peritoneal macrophages of malnourished animals showed reduced functions of adhesion, spreading, and fungicidal activity; production of hydrogen peroxide and nitric oxide were lower, reflecting changes in the innate immune response. However, the high-fat animals had macrophage functions slightly increased. Animals subjected to low-protein diet have immunosuppression, and animals subjected to high-fat diet increased visceral adipose tissue and the presence of an inflammatory process with increased peritoneal macrophage activity and similar systemic changes to metabolic syndrome.


Background
A few decades ago, it was considered unlikely a pandemic of obesity, but from the 1970s, the number of obese people has increased to become worrisome in the 1990s [1]. FAO and WHO data [2,3] also show that the protein malnutrition was a major nutritional disorder throughout the world. However, there has been a rapid change in nutritional status of the population; it characterizes the nutritional transition [4,5]. Changes in diet can lead to significant changes in metabolism, such as the development of various diseases due to the lack of quality in the immune response.
Data from our laboratory and other authors show that among the alterations in malnutrition are the changes in hematopoietic tissue with modification in extracellular matrix components [6,7]. The literature describes alterations in the cell cycle of hematopoietic progenitor [8], reducing the production of cells and hematopoietic precursors [6], both in the erythrocyte series, as in the leukocytes series, leading to anemia and leukopenia [9][10][11]. Protein malnutrition alters both specific and nonspecific immune response of the individual, against infectious agents [11,12].
Obesity is associated with increased levels of leptin and adipocytes, the latter being the main producer of leptin, as well as other mediators, mainly inflammatory, such as tumor necrosis factor alpha (TNFα), IL-1β, and IL-6 that may act directly or indirectly in hematopoietic cells [13]. Leptin also acts on the immune system [14], stimulating the production of white blood cells (WBC) in the bone marrow and their migration. It also increases the production of pro-inflammatory cytokines such as TNFα, IL-6, and IL-12, macrophage adhesion and phagocytosis and stimulates the proliferation of T cells, leading to increased immune competence [15]. Recent studies have shown that obesity leads to decreased blood flow in adipose tissue, causing a hypoxia, which initiates a pro-inflammatory process [16,17], with increased production of acute phase proteins such as C-reactive protein [18] and cytokines such as TNFα and IL-6 [19]. The production of pro-inflammatory cytokines commonly observed in situations of obesity may have effects on hematopoiesis, inducing or inhibiting the proliferation and differentiation of cells present in bone marrow. Inflammatory mediators also have the ability to activate other cells to produce different mediators that perhaps might influence hematopoietic complex mechanism types [16,17]. Some effects exert proliferation (such as GM-CSF), while others have apparently hematopoietic inhibitory effect on proliferation (TNFα and IFNγ, for example) [20,21].
Previous work from our group showed some similarities in immunological parameters between the metabolic condition of malnourished and high-fat fed. We found decreased peritoneal, bone marrow, and serum leukocyte cellularity, as well as reduced production of TNFα and activation of transcription factor NF-kB both in the lowprotein diet [10,22] and in the high-fat diet [23].
The aim of this work is to investigate the similarities and differences between these two diets in macrophage function.

Animals and treatments
Swiss Webster mice, males, 2-3 months old, were obtained in the Faculty of Pharmaceutical Sciences, University of São Paulo. The research project was approved by the Ethics Committee on the Use of Animals of University of São Paulo (protocol: CEUA/FCF/366). The control, low-protein, and high-fat diets (Table 1) were prepared in our laboratory and stored at −4°C until the use. Control diet was prepared for adult mice, according to the recommendations of the American Institute of Nutrition (AIN-93M) [23,24]. The low-protein diet has 2 % of protein [22]. The high-fat diet has 30 % fat in the form of lard, and all other ingredients have been increased by 33 % to provide a balanced nutrient proportion compared to control because, in a pilot project, we found that feed intake in animals subjected in the fat diet was lower than that in the control on this ratio [25]. The animals were housed individually in cages under similar environmental conditions, with clear light/dark cycle of 12/12 h, temperature of 22 ± 2°C, and relative humidity of 55 ± 10 %. The adjustment period to cages was 10-15 days (until the body weight had stabilized). In this period, the animals received the control diet and water ad libitum and were weighed every 48 h to evaluate body weight. After the adaptation, the animals were randomly separated into three groups: control (C), malnourished (M), and high-fat (H). The control group continued to receive the control diet while malnourished began receiving low-protein diet and high-fat group receiving the high-fat diet. During the experiment, body weight and feed were evaluated every 48 h. At the end of 5 weeks, the animals were euthanized for collection of biological material.

Measurement of visceral fat and Lee index
Subsequent euthanasia, the retroperitoneal and periepididymal adipose pads were dissected and immediately weighed. We carry the body mass index (BMI) by Lee index [26], where we use the following link: (weight (g 0.33 )/ naso-anal measured (mm)). These values may characterize the change in mass of mice after the induction of obesity.

Cell culture
Peritoneal cells were obtained by washing the peritoneal cavity with 5 mL of sterile, pyrogen-free McCoy's 5A medium (pH 7.4) supplemented with 10 % fetal bovine serum, glutamine 2 mmol/L, 100 U/mL penicillin, and 100 mg/mL streptomycin (Cultilab, Brazil). Cells were spun down (1500 rpm for 10 min at 4°C) and resuspended twice in McCoy's 5A culture medium. Cell viability was determined by trypan blue exclusion. Cultures rich in macrophages were obtained by incubating 1 × 10 6 cells/mL in 24-or 96-well polystyrene culture plates for 2 h at 37°C in a 5 % CO 2 -humidified air environment. Ingredients have been increased by 33 % to provide a balanced nutrient because that feed intake in animals that subjected the high-fat diet was lower than that in the control on this ratio [25] Non-adherent cells were removed by vigorous washing three times with McCoy's 5A medium. Macrophages were incubated with or without LPS 1 mg/mL (Escherichia coli, serotype 055:B5, Sigma Chemical Company). After 2 or 24 h of incubation with or without LPS, the supernatant was used to determine adhesion, spreading, phagocytosis, and killing, and hydrogen peroxide (H 2 O 2 ) is an oxide nitric (NO) concentrations in the culture supernatant. The entire procedure was executed under aseptic conditions, and all materials used were sterile and free of pyrogen.

Phagocytic and fungicidal activity of peritoneal macrophages
A suspension of C. albicans ATCC Y-837 obtained from a 12-h culture in Sabouraud agar (Difco®) was opsonized with homologous serum obtained from normal mice. Yeast cells were counted in a Neubauer chamber and viability was evaluated using 1 % methylene blue. Only yeast suspensions with over 95 % viability were used. The suspensions were adjusted to 2 × 10 7 yeast cells/mL. In sterile plastic tubes, 500 μl (2 × 10 6 /mL) of the peritoneal cells suspension and 500 μl of the opsonized C. albicans solution were added maintaining a proportion of 1 cell is to 10 yeast cells. The tubes were incubated at 37°C, under agitation at 10 rpm, and aliquots were taken after 30, 60, 90, and 120 min and were cytocentrifuged (Incibras®, Brazil). The coverslips were immediately fixed and stained by May-Grunwald-Giemsa solution. The control of the reaction consisted of tubes containing 500 μl of PBS Dulbecco and 500 μl of the peritoneal cells suspension. For the evaluation of phagocytosis, at least 200 peritoneal cells were counted and those which presented one or more internalized C. albicans cells were considered as having phagocytic activity, the values being expressed in percentage. The fungicidal activity was evaluated accordingly to the technique described by Herscowitz [27] and modified by us.
In this technique, live yeast cells are colored blue by the May-Grunwald-Giemsa stain while the dead yeast cells are not colored at all. The number of yeast cells phagocytosed is variable from cell to cell so the fungicidal activity was expressed by scoring. When the macrophages presented one C. albicans dead, we multiplied by one (score 1), two C. albicans dead multiplied by two (score 2), three C. albicans dead multiplied by three (score 3) and more than four C. albicans dead multiplied by four (score 4). The fungicidal activity was evaluated counting at least 200 macrophages that had phagocytosed C. albicans multiplied for the score according the classification. All samples were processed in duplicate.

NO and H 2 O 2 determination in culture supernatants
Nitric oxide (NO) production was determined according to the Griess colorimetric method [28]. The determination of hydrogen peroxide (H 2 O 2 ) was performed by the method of peroxidase-dependent oxidation of phenol red adapted to micro-assay for PICK & MIZEL [29].

Statistical analysis
Results were checked for distribution normality and homoscedasticity. Means from the C, M, and H groups were compared by unpaired Student's t test or an equivalent non-parametric test (Mann-Whitney). The results from the assays using peritoneal macrophages were analyzed with one-way analysis of variance, Tukey post test, or an equivalent non-parametric test (Kruskal-Wallis, Dunn post hoc). Statistical analyses were performed using GraphPad Prism 5.01 (GraphPad Software, Inc., USA), and the level of significance adopted was 0.05.

Food intake and weight gain
Feed intake was similar between the control and malnourished; however, the consumption of high-fat group was 33 % lower, as already shown in previous experiments in our laboratory. Nevertheless, the calorie intake was similar in all groups because the high-fat group has a higher proportion of lipids in the diet ( Table 2).
In the beginning of the experiment all groups had similar body weight and consumed similar amount of calories for the entire period, the composition of the feed, by varying the amount of lipids and proteins, led high fat group to increase their weight and malnourished animals to significantly decrease body weight.
The animals subjected to high-fat diet had a higher Lee index in comparison to the control group and the malnourished group (p < 0.05). The total weight of the fat pads showed significant differences in the HL group compared to other groups, especially the epididymal fat (Table 2). However, malnourished and control groups showed no significant differences.

Phenotypic characterization of cells in the peritoneal cavity
Phenotypic evaluation of the peritoneal cavity population was done by flow cytometry using the CD11b (FITC) and F4/80 (APC). The double-labeled cells were classified as macrophages. The proportion of macrophages was shown to be decreased in malnourished animals and similar between control and high-fat animals (Fig. 1).

Adhesion and spreading of peritoneal macrophages
The peritoneal macrophages were cultured for 2 h, and after this period, it was shown that cells of malnourished animals had a reduction in macrophage adherence, whereas cells of high-fat animals had increased adhesion. In addition, the cells responded to all three groups were stimulated with LPS. However, cells cultured for 24 h decreased the response to stimulation by LPS. Malnourished animals continue with reduced adhesion, and the control group and high-fat have similar adhesion capacity (Table 3).
The high-fat group presented a lower amount of sprawling cells after 2 h, at baseline and in the group stimulated with LPS compared to control group. The malnourished group showed even less sprawling cells that high-fat even after the stimulus with LPS. Table 3 shows the reaction to the stimulus, demonstrating that the control and high-fat groups respond to the stimulus, and the cells try to spread in larger quantities. The malnourished group did not respond to LPS stimulation. We also observed the spreading (Fig. 2) capacity after 24 h, where we observe that the cells are much spreading, however, remained high-fat and malnourished groups with a lower percentage of control. Analyzing the response to LPS stimulation, we note that the control and high-fat groups continue to react to the stimulus and the malnourished group showed no significant changes.

Phagocytic and fungicidal activity of peritoneal macrophages
The high-fat group showed an increase of phagocytosis of C. albicans after 120 min. However, the malnourished group showed a decrease in phagocytic activity after 90 min, as shown in Table 4.
After the analysis of the opsonized and non-opsonized groups, comparing the control, malnourished, and highfat groups, we note that at 30, 60, and 90 min, the malnourished group had reduced phagocytosis, and the control and high-fat groups were not significantly different. However, at 120 min, there is a similarity in the phagocytic activity between the three groups analyzed.
The fungicidal activity (Fig. 3) was shown to be decreased in the peritoneal cells of malnourished animals at all times evaluated, while it increased in animal cells submitted to high fat diet after 90 min of assay.

Evaluation of hydrogen peroxide and nitric oxide production by peritoneal macrophages
The malnourished group did not react to the stimulus with PMA. Although the groups did not present similar Indicates when there was significant difference between the malnourished and control groups (p < 0.05) stimulated production of hydrogen peroxide, Table 5 shows that the groups have stimulated representative differences and reduced production in malnourished and higher in high-fat group. In unstimulated conditions cells from all groups did not show statistical difference, but when they are stimulated with PMA cells from the control and high-fat group, present increased production of hydrogen peroxide when compared to the unistimulated. The control and high-fat groups responded to LPS stimulation and produced more nitric oxide than unstimulated. The malnourished group did not react to the LPS stimulus. In unstimulated conditions, the cells of animals of all groups showed no statistical difference between them. Although the groups did not present similar stimulated nitric oxide production, Table 5 clearly shows that the groups have stimulated representative differences and greatly reduced production in malnourished and higher in group high-fat group.

Discussion
Changes in dietary components can influence important defense mechanisms against pathogens. We compared the effect of low-protein diet and high-fat diet, on some aspects of innate immunity in mice. Results represent means ± SD of the adherence and percentage of spread cells of peritoneal exudate macrophages of control, malnourished, and high-fat animals (n = 18/group) a Indicates when there was significant difference between the high-fat and control groups (p < 0.05) b Indicates when there was significant difference between the malnourished and control groups (p < 0.05) The animals subjected to high-fat diet showed an increased body weight and macrophage activity with a pro-inflammatory state, similar to metabolic syndrome. However, animals subjected to low-protein diet had a weight reduction with impaired macrophage immune response. Recent evidence suggests that the feed rich in saturated fatty acids induces an inflammatory process in specific regions of the hypothalamus which promotes negative cross-talk with the signaling pathways of leptin and insulin. As a result, it develops a central state of resistance to the action of these hormones, damaging its anorectic effect [30,31].
In this work, it was found that the average feed intake was lower in the group that consumed HFD compared to the control groups. On the other hand, due to the energy density of the HFD be higher than the control diet, caloric intake did not differ between the three groups. Importantly, while consumption of the groups have been isocaloric, we noted the effect of HFD feed on several parameters, such as adiposity, resistance to action of insulin, and the inflammatory response of peritoneal macrophages. This reinforces that the diet composition has a direct influence on health, regardless of energy consumption. Extrapolation of these data to humans suggests that strategies to reduce the risk of obesity and metabolic syndrome cannot be limited to the restriction of calorie intake and should also cover the quality of the diet [23].
The adhesion of macrophages was impaired by malnutrition. The literature reports that the decrease of specific proteins, like fibronectin [32] and integrin, can be the cause of this decrease. The adhesion of macrophages, identified in high-fat animals, can be explained by the action of leptin stimulates adhesion [15] in addition to the production of inflammatory cytokines that stimulate macrophage activity [33]. Works with malnourished humans show decreased adhesiveness due to changes in the activity of macrophages [34].
Few studies have been done with regard to adhesion and spreading of macrophages in malnourished and obese mice. Some studies of our group showed that in malnourished animals, there is a reduction in the spreading of macrophages [12,35]. In our study, we observed a decrease in the spreading of macrophages in malnourished animals, both in time of 2 h for 24 h. It was also observed that stimulation with LPS had no effect on any of the times evaluated in cells derived from malnourished animals. This reduction in activity may be linked to the impairment of macrophage activity, characterized by decreased production of cytokines stimulating the activation of macrophages, such as IL-6 and TNF [10,16].  Results represent means ± SD of the percentage of phagocytic and fungicidal activity of peritoneal exsudate macrophages of control, malnourished and high-fat animals (n = 15/group) a Indicates when there was significant difference between the high-fat and control groups (p < 0.05) b Indicates when there was significant difference between the malnourished and control groups (p < 0.05) Phagocytosis is a process of particle aggregation in excess of 1 μm. Its main objective is death and/or inactivation of pathogens [36]. The decrease in phagocytosis of cells in malnourished animals can be related to the decreased production of pro-inflammatory cytokines that stimulate macrophage activation. The increase phagocytosis in cells derived from animals in the highfat group may be due to the amplification of the production of pro-inflammatory cytokines that stimulate macrophage activation [37]. The phagocytic processes are driven by a finely controlled rearrangement of the cytoskeleton. A variety of signals may converge to rearrange the actin cytoskeleton in a phagosome. Studies have shown the complexity of the phagocytic signals, such as the involvement of lipid complexes and signaling in the transduction of signals from the cytoskeleton phagocytic receptors [38,39].
The fungicidal activity is decreased in the malnourished group. This may be due to decreased production of reactive species of oxygen and nitrogen, essential for this process [40,41]. The high-fat group showed an increase in fungicidal activity. As we have increased the production of reactive species of oxygen and nitrogen in high-fat animals, this can enhance the fungicide process.
Nitric oxide (NO) is a basic nitrogen reactive species to various metabolic activities such as vasorelaxation, macrophage-mediated cytotoxicity, activation inhibition, platelet adhesion and aggregation, regulation of basal and blood pressure. Studies have shown that proteins or specific nutrient deficiency leads to the reduction of nitric oxide [41,42]. Our study showed that the malnourished animals had decreased nitric oxide production. This result may be due to the lower availability of arginine, basic substrate for the production of nitric oxide [43] or changes in iNOS enzyme system [44]. In animals treated with high-fat diet, we found increased production of nitric oxide, as the largest macrophage activation resulting from stimuli such as MCP-1 [33,45].
Hydrogen peroxide (H 2 O 2 ) is a key reactive oxygen species in the lysis of phagocytized microorganisms which oxidize the cell membrane to form disulfide bridges between the cysteine of different structural proteins of the bacterium, leading to the death of the same [46]. In our study, we found a lower production of hydrogen peroxide in cultures of 2 h of peritoneal macrophages of malnourished animals [40]. By stimulating macrophages in culture with PMA (phorbol-myristate-acetate), we note that the malnourished group was the only one not to respond to the stimulus. This may be a result of protein deficiency that causes changes in the enzyme system that triggers oxidative stress through NADPH oxidase [44].  Results represent means ± SD of the production of hydrogen peroxide and nitric oxide by peritoneal exsudate macrophages of control, malnourished and high-fat animals (n = 18/group) a Indicates when there was significant difference between the high-fat and control groups (p < 0.05) b Indicates when there was significant difference between the malnourished and control groups (p < 0.05) Regarding high-fat animals, we found a greater production of hydrogen peroxide (after 2 h) in macrophage culture supernatant, which may be explained by the increased activation of macrophages by pro-inflammatory cytokines that stimulate phagocytic activity, bactericidal, and fungicidal, leading to an increase of reactive oxygen species [47].
The data obtained throughout the study and the realization of this work showed malnourished animals with compromised immune response. The animals submitted to high-fat diet showed increased macrophage activity with inflammation similar to metabolic syndrome. These data are consistent with the literature.

Conclusions
The protein malnutrition impairs the innate immune response by decreasing the amount and functionality of macrophages, which are poorly responding to the activation and production of substances. In the murine model of high-fat diet, we observed an increase in visceral adipose tissue and the presence of an inflammatory process with increasing activity of macrophages in that tissue and systemic changes similar to the metabolic syndrome.