Visceral adipose tissue (VAT) is considered to be an energy storage depot and has now been recognized as a highly active organ responsible for controlling numerous metabolic, hormonal, and immune processes [1
]. Moreover, it is well-accepted that a large VAT mass is a main risk factor for developing metabolic dysfunction [4
]. This tissue is constituted mainly by two components: (1) mature adipocytes that are responsible for controlling metabolism by storing excess calories as lipids as well as the production of endocrine hormones, such as leptin, resistin, and adiponectin [5
]; and (2) stromal vascular fraction (SVF), which is a rich source of preadipocytes, endothelial cells, fibroblasts, mesenchymal cells, hematopoietic progenitor cells, and immune cells [6
]. Of the heterogeneous leukocytes in VAT, macrophages are considered to be important contributors to adipose tissue maintenance, regulating immune and metabolic functions through cross-talk with adipocytes [7
It has been demonstrated in animal models and human adipose tissue that macrophage phenotype is shaped by the microenvironment in the adipose tissue depots. Adipose tissue macrophage (ATM) accumulation and polarization from the anti-inflammatory (M2) to the pro-inflammatory (M1) phenotype have been associated with the metabolic and inflammatory changes that are the hallmarks of obesity [2
On the other hand, the growing rates of overweight women at reproductive age and the increasing number of overweight or obese women beginning pregnancy are considered to be serious public health issues in the 21st century [9
]. This has consequences in the mother and has subsequent effects on the child, leading to obesity, dyslipidemias, Type 2 diabetes, and hypertension later in life [10
The potential role of ATMs in pregnancy as regulators of maternal inflammation is not yet fully understood. Therefore, the aim of this study was to characterize macrophage populations in VAT from pregnant women and to evaluate differences by weight status.
2. Materials and Methods
2.1. Patients and Donor Selection
Protocol 212250-3210-21002-06-15 was approved by the IRB of the Instituto Nacional de Perinatologia in Mexico City. Each participant signed a written informed consent form.
A cross-sectional study was conducted from January 2016 to December 2016 at Instituto Nacional de Perinatologia, Mexico City, Mexico. Women were recruited consecutively and written informed consent was obtained from each one. During this period, 1294 newborns were delivered by cesarean section at term (37–40 weeks of gestation by obstetric ultrasonography) in our institution. A total of 69 women met the inclusion criteria, 34 declined to participate in the study, it was not possible to obtain samples from 8 women, and 6 collected samples were insufficient for the experiments. Twenty-one healthy pregnant women with intact membranes, no clinical evidence of intrauterine infection, and normal weight gain during pregnancy were enrolled according to the inclusion criteria: age 20–40 years, pregestational Body Mass Index (BMI) 18.5 kg/m2 and known pregestational weight. Exclusion criteria included diagnosis of gestational or pre-gestational diabetes, thyroid, hypertensive, or immunological disease, treatment with drugs that could affect metabolism or inflammation during pregnancy, intrauterine growth restriction, or fetal genetic abnormalities. Maternal weight and height at the end of pregnancy were measured by trained personnel. Pregestational weight was collected from clinical files and BMI was calculated according to the World Health Organization criteria. The subjects were stratified into three groups: normal weight (BMI 18.5–24.9 kg/m2; n = 7), overweight (BMI 25.0–29.9 kg/m2; n = 7), and obesity (BMI 29.9 kg/m2; n = 7).
2.2. Collection of Visceral Adipose Tissue
Visceral Adipose Tissue biopsies (5 × 5 cm) were obtained from the omentum (middle bottom portion) prior to the hysterotomy for delivery under sterile conditions. Samples were transported in sterile recipients and processed up to 1 h after collection.
2.3. Isolation of the Stromal Vascular Fraction from Visceral Adipose Tissue
Visceral Adipose Tissue biopsies were washed in 1× Phosphate Buffered Saline (1× PBS) before being minced with sterile scissors to be digested with collagenase solution (0.01 M PBS, 5 mM glucose, 1.5% w/v BSA, and 0.25% collagenase type I (GIBCO by Life Technologies, Carlsbad, CA, USA) at a pH of 7.4) for 1 h at 37 °C under constant shaking. After this, the digested tissue was filtered and centrifuged at room temperature (RT) for 10 min at 200× g to separate adipocytes by flotation from the stromal vascular fraction. After discarding the supernatant, the cell pellet containing preadipocytes, endothelial cells, and immune cells was incubated in erythrocyte lysis buffer (Roche, Basel, Switzerland) for 5 min at RT. The reaction was stopped by addition of ice-cold 1× PBS before the solution was centrifuged at 4 °C for another 5 min at 1500× g. The cell pellet was washed with 1× PBS twice under the same conditions, and cells were counted in a standard hemocytometer using trypan blue staining to assess viability.
2.4. Immunostaining for Macrophages from Stromal Vascular Fraction
The stromal vascular fraction cell suspensions (1 × 106 cells/mL) were incubated with the labelled monoclonal antibodies CD45/PE-Cy7 (Cat. 304016; clone HI30), CD14/APC-Cy7 (Cat. 325620; clone HCD14), HLA DR/PE-Cy5 (Cat. 307608; clone L243), CD11c/PE (Cat. 301606; clone 3.9), CD163/FITC (Cat. 333618; clone GHI/61), CD206/APC (Cat. 321110; clone 15-2), and DAPI solution (Cat. 422801) for 15 min at RT in the dark. All antibodies were purchased from BioLegend (San Diego, CA, USA). Appropriate Fluorescent Menus One controls were also used. Cells were fixed using a fluorescence-activated cell sorting (FACS) lysing solution (Cat. 349202, BD Bioscience, Frankling Lakes, NJ, USA) according to the manufacturer’s instructions. Flow cytometry analysis was performed using a FACS Aria III with the FASCDiva software 6.0 (BD Bioscience).
The gating strategy used to identify macrophage subpopulations in the stromal vascular fraction consisted of excluding aggregated and dead cells. After this, cells were gated based on CD45 versus CD14 expression as either CD45−CD14+ (resident macrophages) or CD45+CD14+ (recruited macrophages). The HLA-DR, CD11c, CD163, and CD206 expression was determined as the median fluorescence intensity (MFI) for each marker.
2.5. Hematoxylin and eosin stain of Adipose Tissue Sections
Paraffin-embedded 5-µm VAT sections fixed in 10% formaldehyde were deparaffinized, rehydrated, stained with hematoxylin and eosin, and mounted using a DPX mounting medium (Sigma-Aldrich, St. Louis, MO, USA). A minimum of four randomly chosen low-power (20× fields were counted by the same observer to determine adipocyte number and size per field in VAT using the AxioVision version 4.9.1 Microscopy Software (Zeiss, Oberkochen, Germany).
2.6. Statistical Analysis
Distribution analysis was assessed using the Shapiro–Wilk normality test. Multiple groups were analyzed by one-way ANOVA or Kruskal–Wallis one-way analysis of variance, while pairs of groups were compared by the Mann–Whitney U test as appropriate for variables that were not normally distributed. The relationship between two variables was determined by Spearman’s rank correlation coefficient. A linear regression model was performed to evaluate the effect of BMI on the variables of interest. Statistical analyses were performed with the IBM (Armonk, NY, USA) SPSS 20.0 statistical software. Unless otherwise stated, all data were expressed as mean ± SD (standard deviation) or median and interquartile range. In the present study, p 0.05 was considered statistically significant.
4. Discussion and Conclusions
In this study, we have shown the presence of CD45−CD14+ macrophage cells in VAT obtained from pregnant women for the first time. Furthermore, we demonstrated that being overweight and obese in the pregestational period is associated with adipocyte hypertrophy and specific ATM populations. The resident HLA-DRlow/− macrophage number is significantly decreased in women with pregestational obesity compared to normal weight. On the other hand, we described a positive correlation between CD206 expression and BMI in all macrophage subsets as well as an increased CD11c expression associated with obesity in recruited HLA-DR+ macrophages.
Analysis of markers CD45 and CD14 used to identify monocyte/macrophage linage cells in VAT revealed the typical CD45+
population with a hematopoietic origin and a newly described CD45−
cell population that we propose to be cells with a non-hematopoietic origin. These findings are similar to other reports of non-hematopoietic origin macrophages in different tissues. [12
]. Moreover, we were able to define two subsets of recruited macrophages (CD45+
) that originate from circulating monocytes and are best known as monocyte-derived macrophages. We also defined two subsets of CD45−
HLA-DR) that may be those described as resident macrophages in adipose tissue [17
]. Our results support the statement that these cells are independent of hematopoietic linage and are crucial for homeostasis and tissue immunity in several tissues, including adipose tissue [19
There are several hypotheses about the origin of resident macrophages. Hoeffel and Ginhoux proposed that resident macrophages may derive from hematopoietic stem-cell-independent embryonic precursors [20
]. Other authors have concluded from in vitro animal models that these cells could have come from a population of preadipocytes with a macrophage-like phenotype [21
] or that they have developed in adipose tissue from preadipocytes given their macrophage-like features, such as phagocytic activity, microbicide activity, and expression of some macrophage antigens [22
We observed differences in the distribution of resident and recruited macrophages among the study groups. Tissue-resident macrophages have a role in the maintenance of white adipose tissue, acting as a link between adipocytes and immune cells to coordinate tissue remodeling and function of white adipocytes [23
]. As a consequence, the lower number of resident macrophages found in women with pregestational obesity compared to normal weight in our study could be associated with a loss in tissue maintenance and energy homeostasis in VAT.
Several studies have shown that the HLA-DR marker is associated with immunoregulatory functions in different pathologies, suggesting that antigen DR expression reflects the macrophage activation degree. A low DR expression indicates suppressed macrophage activation [24
], while a high DR expression reflects activation of the inflammatory NF-KB pathway [25
]. Based on this evidence, it seems reasonable that macrophage populations expressing HLA-DR+
are inflammatory and non-inflammatory macrophages, respectively.
Some authors propose that the imbalance towards an inflammatory environment in obese adipose tissue may result from an accumulation of pro-inflammatory macrophages that originated either locally or were recruited from the circulation [26
]. In our study, the lower number of resident HLA-DRlow/−
macrophages found in women with pregestational obesity compared to normal women could reflect a loss of adipose tissue homeostasis, which favors a pro-inflammatory environment. Additionally, the lower number of resident HLA-DRlow/−
macrophages strongly correlates with a gradual increase in pregestational BMI, which supports our hypothesis.
Although we also found significant differences in the recruited populations among the study groups, our results suggest that pregestational BMI has a bigger influence on the number of resident cells. Therefore, the changes observed in obese adipose tissue may be due to the decreased number of cells with regulatory functions. Despite resident HLA-DRlow/− macrophages possibly having anti-inflammatory properties, our study design did not allow us to fully determine these characteristics and ongoing studies in our laboratory are focused on further characterizing macrophage population functionality in VAT.
The macrophage phenotype has been described as a spectrum from classically activated macrophages (M1) to alternatively activated macrophages (M2). Studies in mice models showed that early stages of adipose tissue expansion are characterized by M2-polarized macrophages with progressive lipid accumulation promoting M1 polarization [30
]. CD11c is an integrin used as a monocyte/M1 macrophage activation marker. This marker is increased in peripheral monocytes from patients with hyperlipidemia, which plays an important role in the development of atherosclerosis. Some reports have focused on the role of CD11c in adipose tissue, showing that pro-inflammatory adipose tissue macrophages expressing high CD11c levels are associated with inflammation and insulin resistance [31
]. In our study, women with a high pregestational BMI (≥30) expressed elevated CD11c levels in resident HLA-DR+
macrophages, suggesting that the inflammatory profile observed in obese pregnant women is elicited mainly from resident cells.
On the other hand, alternative macrophages that classically express high CD206 levels have been inversely associated with metabolic disturbances and seem to be present mainly in healthy adipose tissue expansion [33
]. A subpopulation expressing CD206+
was associated with obesity and insulin resistance [31
]. Interestingly, we found that both resident and recruited macrophages in VAT from women with high BMI expressed high CD206 levels in our study, particularly the recruited HLA-DR+
macrophage subpopulation from women with pregestational obesity. This has not been previously reported.
Some authors speculate that in order to control adipose tissue inflammation, M1 macrophages switch activation to an anti-inflammatory phenotype. As observed in our study, pro-inflammatory macrophage subsets also express high levels of CD206 and CD163 markers that usually characterize M2 macrophages, which could be consistent with the hypothesis that macrophages could reverse their phenotype and function depending on the microenvironment [34
It is widely known that obesity promotes robust changes in adipose tissue morphology, including adipocyte hypertrophy and hyperplasia, as well as changes in the composition of immunological cells [4
]. This adipocyte hypertrophy has been associated with adipokine production, cell death, and local adipose hypoxia that contributes to macrophage accumulation and promotes inflammation in adipose tissue [35
]. However, most of these studies have been performed in animal models using an obesogenic diet or in human morbid obesity, overlooking the impact of maternal obesity on human reproductive health and early development [36
In the present study, we describe the morphologic features of adipocytes present in VAT from pregnant women. We were able to determine a direct correlation with changes in size and number depending on pregestational weight. Our results show that VAT adipocyte hypertrophy and reduction in the number of these cells at the end of pregnancy correlate with being overweight or obese in the pregestational period. It is important to note that in the overweight state, there is moderated hypertrophy as the severity of this condition increases with BMI. Adipocyte hypertrophy may be associated with a higher degree of inflammation in these groups of women. This may lead to adverse perinatal outcomes, such as glucose intolerance and insulin resistance, as has been demonstrated in previous reports using pregnant mice models [42
]. Our findings are similar to those reported by Haghiac et al., which described that adipocytes from obese pregnant women are larger than those from lean pregnant women, although these authors analyzed abdominal subcutaneous fat [44
]. Regarding animal models, Zhang et al. 2011 showed that adipocytes obtained from subcutaneous and parametrial adipose tissue in lean pregnant mice are larger than those from lean non-pregnant mice, concluding that adipocyte hypertrophy is dependent on pregnancy [45
]. The strong correlation of BMI with adipocyte size and number found in our study suggests that hypertrophy depends on weight rather than pregnancy state, even though we did not evaluate non-pregnant women.
Human VAT is constituted by adipocyte and non-adipocyte cells, such as endothelial cells, fibroblasts, preadipocytes, and several immune cells, such as macrophages [46
]. It has been demonstrated that cross-talk is established between them and regulates tissue functionality at different levels [47
]. Moreover, macrophages seem to play a leading role in the inflammatory response elicited by this tissue in obesity [6
]. Some studies have characterized macrophage populations and composition using different sources of adipose tissue, which is mostly related to extreme degrees of obesity. This showed that a high infiltration of pro-inflammatory macrophages, especially in VAT, strongly correlates with a local and systemic inflammatory environment [50
]. Haghiac et al. reported a higher number of macrophages in subcutaneous adipose tissue from obese pregnant women compared to lean pregnant women, which was correlated with an inflammatory state and higher release of cell-free DNA in the obese group [44
]. However, this study lacks a characterization of macrophage populations and does not include an overweight women group as a previous state to obesity. All findings are summarized in Figure 5
It is important to consider that this macrophage characterization was conducted only at the resolution of pregnancy due to tissue availability and it is not possible to establish if pregnancy itself has any influence on the changes in macrophage subsets. In addition, our experimental design did not allow us to determine the origin and functionality of resident macrophages in human VAT, so further research should focus on this premise.
Overall, these novel findings suggest that adipose tissue inflammation in women that are overweight or obese in the pregestational period is strongly associated with changes in the proportion of the macrophage subpopulations and differences in the expression of anti- and pro-inflammatory surface markers in these cells. According to our results, the first changes could have occurred in the resident macrophages in these two groups of patients, which decrease the number and expression of higher levels of pro- and anti-inflammatory markers. Recruited cells also seem to show differences, particularly the CD45+CD14+HLA-DR+ subset that expresses high levels of anti-inflammatory markers as BMI increases. We infer that these events may occur in order to compensate for the inflammatory process and maintain tissue homeostasis. Leading with this, it would be interesting to explore whether tissue inflammation is more severe in women with excessive weight gain during pregnancy as well as characterize the composition and distribution of macrophage subsets and their potential association with the onset of metabolic diseases during pregnancy and later in the mother and offspring.