Introduction

Adipose tissue dysfunction is a hallmark of obesity-associated metabolic disturbances. It is characterised by a significant reduction in its capacity to store excess fuel [1, 2] and to generate new adipocytes (adipogenesis) [3, 4] in parallel with increased levels of inflammatory markers, as a consequence of macrophage infiltration [5] and adipocyte inflammation [6].

Recent studies demonstrated that iron excess in adipose tissue promotes adipose tissue dysfunction, leading to decreased adipogenesis and enhanced adipocyte inflammation and adipose tissue macrophage M1 polarisation [7,8,9,10,11,12]. Intracellular mediators of iron homeostasis contribute to changes in adipose tissue physiology and adipogenesis, including SLC40A1 or ferroportin [9], transferrin [11], cytosolic aconitase [13] and mitoNEET [14, 15], but less is known about the source of iron excess.

Haem oxygenase 1 (HMOX1, also known as HO-1) is an inducible enzyme required for haem degradation. Since haem degradation leads to the production of biliverdin, ferrous iron and carbon monoxide, haem oxygenase activity is an important source of intracellular iron. Taking into account that high dietary intake of haem iron, which is more easily absorbed than non-haem iron [16], is considered a risk factor in the progression of type 2 diabetes and obesity-associated metabolic disturbances [16,17,18,19], and the importance of intracellular haem levels in adipogenesis [20], we hypothesised that increased adipose tissue HMOX1 activity (an extra source of iron) might contribute to obesity-associated adipose tissue iron excess and, in consequence, promote adipose tissue dysfunction. In line with this hypothesis, a recent study demonstrated that macrophage HMOX1 is required to promote adipose tissue inflammation in mice [21]. Specifically, macrophage HMOX1 deletion improved adipose tissue dysfunction, inhibiting the development of crown-like structures and macrophage infiltration, reducing adipocyte hypertrophy and increasing the expression of adipogenic genes and the number of protective M2 macrophages. Furthermore, this study also showed that the inhibition of HMOX1 activity in liver and in macrophages improved glucose clearance and insulin sensitivity, and decreased high-fat diet (HFD)-induced inflammation [21]. In contrast to this study, two previous studies in murine models demonstrated that systemic induction of HMOX1 activity by intraperitoneal cobalt protoporphyrin administration resulted in reduced adiposity, increased circulating adiponectin and improved insulin sensitivity and glucose tolerance [22, 23].

In humans, a few studies have examined HMOX1 gene expression in adipose tissue. Specifically, Shakeri-Manesch et al reported the first study showing increased subcutaneous (SAT) and visceral (VAT) adipose tissue HMOX1 mRNA and protein levels in obese participants (morbidly obese [n = 20] vs non-obese [n = 20]) [24]. VAT HMOX1 expression was negatively correlated with HOMA-IR [24]. Two additional studies confirmed increased SAT (n = 42 [25]) and VAT (n = 50 [21]) HMOX1 gene expression in obese participants, but in positive association with insulin resistance. In one of these studies the significance of the correlation between SAT HMOX1 mRNA and insulin resistance was lost after controlling for BMI [25]. The studies indicate a clear association between adipose tissue HMOX1 expression and adiposity. However, the association with insulin resistance remains controversial.

To the best of our knowledge the possible role of HMOX1 in the relationship between iron metabolism and adipogenesis in human adipose tissue and adipocytes has not been explored. In the current study, we aimed to investigate the possible role of HMOX1 in human adipose tissue, adipogenesis and iron metabolism. To this end, we evaluated (1) SAT and VAT HMOX1 gene expression in two independent cohorts in relation with obesity; (2) the impact of weight loss and weight gain (in rats) on adipose tissue HMOX1; (3) HMOX1 mRNA levels during human adipocyte differentiation; (4) the effects of inflammation, iron and haem arginate (hemin) on adipocyte HMOX1; and (5) the effects of induction of HMOX1 activity on adipocyte mitochondrial respiratory function, glucose uptake and adipogenesis.

Methods

Patient recruitment

Adipose tissue samples were obtained from three independent cohorts. In cohort 1, a group of 234 adipose tissue samples (120 VAT and 114 SAT) from participants with normal body weight and different degrees of obesity (BMI ranging from 20 to 68 kg/m2) were analysed. In cohort 2, which comprised morbidly obese (BMI >35 kg/m2) participants with different degrees of insulin action (measured using hyperinsulinaemic–euglycaemic clamp [9] as detailed in the electronic supplementary material [ESM] Methods), 40 paired SAT and VAT samples were studied. All these participants were recruited at the Endocrinology Service of the Hospital of Girona ‘Dr Josep Trueta’. All participants gave written informed consent, validated and approved by the Ethics Committee of the Hospital of Girona ‘Dr Josep Trueta’, after the purpose of the study was explained to them.

In cohort 3, 21 obese (47 ± 9 years [mean ± SD]) participants of European descent, who underwent bariatric surgery via Roux-en-Y gastric bypass (RYGB) at the Hospital of Girona ‘Dr Josep Trueta’ were part of an ongoing study [9, 14]. Adipose tissue samples from the SAT depot were obtained during bariatric surgery. Postoperative samples of SAT were obtained by subcutaneous biopsy at the mesogastric level 2 years after surgery. All participants gave written informed consent, validated and approved by the Ethics Committee of the Hospital of Girona ‘Dr Josep Trueta’, after the purpose of the study was explained to them.

Additional information is detailed in ESM Methods.

Adipose tissue handling

Adipose tissue samples were obtained from SAT and VAT (omentum) depots during elective surgical procedures (cholecystectomy, abdominal hernia surgery and gastric bypass surgery). Samples of adipose tissue were immediately transported to the laboratory (5–10 min). The handling of tissue was carried out under strictly aseptic conditions. Adipose tissue samples were washed in PBS, cut with forceps and scalpel into small pieces (100 mg) and immediately flash-frozen in liquid nitrogen before being stored at –80°C. The isolation of adipocyte and stromal vascular fraction (SVF) cells was performed from nine non-frozen SAT samples as described in the ESM Methods.

Effects of weight gain in rats

Four-week-old male Wistar rats (n = 20) were randomised into two groups fed ad libitum during an average of 6 months either a normal chow diet (n = 10) for comparative purposes or an HFD (n = 10) to induce obesity. These experiments are detailed in the ESM Methods. All experimental procedures conformed to the European Guidelines for the Care and Use of Laboratory Animals (directive 2010/63/EU) and were approved by the Ethics Committee for Animal Experimentation of the University of Navarra (049/10). No data from these experiments has been excluded.

Differentiation of human pre-adipocytes

Differentiation of human pre-adipocytes is described in the ESM Methods. For the time course experiment, cells were harvested and stored at –80°C for RNA extraction at day 0, 2, 5, 7, 9, 12 and 14. The effect of administration of (1) macrophage-conditioned medium (MCM, 2.5%) from THP-1 cells previously treated with lipopolysaccharide (10 ng/ml for 24 h), (2) hemin (60 μmol/l), (3) FeSO4 (3 and 30 μg/ml) or (4) deferoxamine (DFO, 20 μmol/l) during subcutaneous adipocyte differentiation was studied. Otherwise, before and after adipocyte differentiation (at day 0 and day 14), subcutaneous pre-adipocytes and fully differentiated adipocytes were incubated with fresh medium (control) or fresh medium containing several treatments. These treatments included hemin (60 μmol/l), FeSO4 (3 and 30 μg/ml) and DFO (20 μmol/l). After 48 and 72 h, cells were harvested, and pellets and supernatants were stored at –80°C. The concentrations of these treatments were selected and optimised according to previous reports [10, 20]. All in vitro experiments were randomised and performed in triplicate. Experimenters were blind to group assignment and outcome assessment. No data from these experiments has been excluded.

Glucose uptake

Adipocyte glucose uptake was measured using the Glucose Uptake Assay Kit (ab136955, Abcam, Cambridge, UK) as described in the ESM Methods.

Mitochondrial respiratory function

Mitochondrial respiratory function was assessed using the Seahorse XFp Extracellular Flux Analyzer (Seahorse Bioscience, Agilent Technologies, Santa Clara, CA, USA) using the Seahorse XFp Cell Mito Stress Test Kit according to the manufacturer’s instructions.

RNA expression

RNA purification, gene expression procedures and analyses were carried out as previously described [9, 14]. Briefly, RNA purification was performed using an RNeasy Lipid Tissue Mini Kit (QIAgen, Izasa, Barcelona, Spain) and the integrity was checked using an Agilent Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). Gene expression was assessed by real-time PCR using a LightCycler 480 Real-Time PCR System (Roche Diagnostics, Barcelona, Spain), using TaqMan and SYBR green technology suitable for relative genetic expression quantification. The commercially available and pre-validated TaqMan primer/probe sets used are detailed in the ESM Methods.

Adipose tissue protein extraction

VAT (n = 40) and SAT (n = 26) protein were extracted as described in the ESM Methods. HMOX1 protein levels were measured using Human HO 1 ELISA kit (ab133064, Abcam) in adipose tissue extracts following manufacturer’s instructions. Total iron was measured using the Iron-FerroZine colorimetric method following the manufacturer’s protocol (11509, Biosystems, Barcelona, Spain). The same amount of total protein (100 μg) was used to perform these measurements.

Analytical methods

Serum glucose, insulin, homeostasis model assessment of insulin resistance (HOMA-IR), HDL-cholesterol, fasting triacylglycerols and C-reactive protein concentration were measured as previously described [9, 14].

Statistical analyses

Statistical analyses were performed using SPSS 12.0 software. Unless otherwise stated, descriptive results of continuous variables are expressed as mean and SD for Gaussian variables or median and interquartile range. Variables that did not fulfill normal distribution criteria were Log10 transformed to improve symmetry for subsequent analyses. The relation between variables was analysed by simple correlation (Pearson’s test and Spearman’s test) and multiple regression analyses in a stepwise manner. One-factor ANOVA with post hoc Bonferroni test, paired t test and unpaired Student’s t test were used to compare HMOX1 gene expression according to obesity and type 2 diabetes. A nonparametric test (Mann–Whitney U test) was used to analyse in vitro experiments. Levels of statistical significance were set at p < 0.05.

Results

HMOX1 mRNA in human adipose tissue

Anthropometric and clinical variables of participants in cohort 1 are shown in Table 1. No significant differences in SAT and VAT HMOX1 expression were found in non-obese (0.019 ± 0.01 vs 0.020 ± 0.02, p = 0.7) and obese participants (0.030 ± 0.01 vs 0.031 ± 0.01, p = 0.6). In both SAT and VAT, HMOX1 gene expression was significantly increased in obese participants (Table 1). Both SAT and VAT HMOX1 gene expression was associated with BMI, but not with fasting glucose (Table 2). In this cohort, the increased SD of fasting glucose in obese compared with non-obese participants (6.2 ± 2.5 vs 5.0 ± 0.5 mmol/l [mean ± SD]) might explain the lack of association between adipose tissue HMOX1 and fasting glucose in these participants. In obese participants, both SAT and VAT HMOX1 gene expression values above the first quartile were significantly associated with increased fasting glucose (ESM Table 1, Fig. 1). VAT HMOX1 gene expression was also positively correlated with fasting triacylglycerols (Table 2). However, this correlation was lost (r = 0.06, p = 0.5) after controlling for BMI.

Table 1 Anthropometric and clinical variables of study participants from cohort 1 and cohort 2
Full size table
Table 2 Bivariate correlation among clinical variables, expression of HMOX1 gene and expression of adipose tissue-related genes in VAT and SAT from cohort 1
Full size table
Fig. 1
figure 1

(a, b) Fasting glucose according to VAT (a) and SAT (b) HMOX1 gene expression first quartile in cohort 1. (c, d) Fasting glucose according to VAT (c) and SAT (d) HMOX1 gene expression first quartile in cohort 2. *p < 0.05 compared with HMOX1 mRNA (RU) <1st quartile. Error bars show 95% CI. In cohort 1, HMOX1 gene expression values were 0.028 (0.021–0.038) RU in SAT and 0.027 (0.019–0.036) RU in VAT (median and interquartile range) and the first quartiles were 0.021 RU in SAT and 0.019 RU in VAT. In cohort 2, HMOX1 gene expression values were 0.033 (0.027–0.048) RU in SAT and 0.031 (0.020–0.042) RU in VAT (median and interquartile range) and the first quartiles were 0.027 RU in SAT and 0.020 RU in VAT

Full size image

HMOX1 gene expression was negatively correlated with adipogenic (PPARG, FASN, ACACA, SLC2A4), insulin pathway-related (IRS1) and mitochondrial biogenesis-related (PPARGC1A [only in SAT], MTCO3) genes, and positively correlated with markers of inflammation (TNF, IL10, IL6, IL8), iron accumulation (FTL, SLC40A1) and oxidative stress (CYBA, NFE2L2) in both SAT and VAT (Table 2). HMOX1 gene expression was negatively correlated with CD206 (also known as MRC1)/CD68 ratio (a specific marker of M2 macrophages) in VAT, but not with expression of the non-specific classical macrophage marker, CD68 (Table 2). In multivariate regression analysis, all these gene expression associations (except IRS1 in SAT and CD206/CD68 in VAT) remained significant after controlling for age and BMI in both VAT (ESM Table 2) and SAT (ESM Table 3).

In a subgroup of participants, HMOX1 protein and total iron levels were measured. Significant associations among HMOX1 mRNA, HMOX1 protein and total iron levels in both VAT (Fig. 2a–c) and SAT (Fig. 2d–f) were found. HMOX1 mRNA, HMOX1 protein and total iron levels were increased in those participants with BMI over 35 kg/m2 (ESM Fig. 1).

Fig. 2
figure 2

Bivariate correlations among HMOX1 gene expression, HMOX1 protein and total iron levels in VAT (a–c) and SAT (d–f). Log indicates Log10 transformation

Full size image

A second independent cohort (cohort 2) composed of morbidly obese patients confirmed the association between HMOX1 gene expression and markers of intracellular iron excess (SLC40A1 [r = 0.41, p = 0.03], FTL [r = 0.46, p = 0.005] and CYBA [r = 0.61, p < 0.0001]) and inflammation (TNF [r = 0.58, p < 0.0001], LBP [r = 0.53, p = 0.001]). The association between both SAT and VAT HMOX1 gene expression values above the first quartile and fasting glucose were also replicated in cohort 2 (ESM Table 1, Fig. 1). SAT HMOX1 gene expression values above the first quartile were also associated with HOMA-IR (ESM Table 1). In this cohort, HMOX1 gene expression was non-significantly increased in SAT vs VAT (0.041 ± 0.019 vs 0.032 ± 0.018 p = 0.06).

Effects of bariatric surgery-induced weight loss

In cohort 3, bariatric surgery-induced weight loss resulted in decreased SAT HMOX1 mRNA levels (Table 3). The per cent change in HMOX1 mRNA levels was positively correlated with the per cent change in weight (r = 0.51, p = 0.01), BMI (r = 0.60, p = 0.004) and fat mass (r = 0.66, p = 0.001), but not with the per cent change in fasting glucose, HDL-cholesterol, fasting triacylglycerols, C-reactive protein or insulin sensitivity (estimated using M value) (Table 3). In addition, the per cent change in HMOX1 mRNA levels was negatively correlated with the per cent change in adipogenic (ADIPOQ, SLC2A4) gene expression and positively correlated with the per cent change in intracellular iron excess-related (SLC40A1, FTL) gene expression (Table 3). No correlation was found between the per cent change in HMOX1 and inflammatory-related (TNF, IL6, IL8, CD14) gene expression (Table 3).

Table 3 Effects of bariatric-surgery-induced weight loss on HMOX1 gene expression, anthropometric and clinical variables and other gene expression in SAT in cohort 3
Full size table

Effects of HFD-induced weight gain in rats

HFD-induced weight gain resulted in increased Hmox1 mRNA levels in epididymal white adipose tissue (eWAT), positively correlated with weight, eWAT weight, HOMA-IR, circulating leptin and eWAT Il6, and negatively correlated with Adipoq mRNA levels (Fig. 3a–i).

Fig. 3
figure 3

(a–c) Effects of HFD in Hmox1 mRNA (a), weight (b) and eWAT weight (c) in rats. **p < 0.01, ***p < 0.001 compared with normal diet. Error bars show 95% CI. (d–i) Bivariate correlation between eWAT Hmox1 mRNA and weight (d), eWAT weight (e), HOMA-IR (f), circulating leptin (g), and eWAT Il6 (h) and Adipoq (i) mRNA

Full size image

HMOX1 in human adipogenesis

HMOX1 mRNA levels progressively decreased during differentiation of human adipocytes (Fig. 4a). The analysis of adipose tissue subfractions showed that HMOX1 mRNA levels were significantly reduced in adipocytes compared with SVF cells (including CD14+ and CD14− cells) (Fig. 4b), with the highest expression being shown in CD14+ cells (enriched with adipose tissue macrophages) (Fig. 4b). Inflammatory conditions (MCM, 2.5%; Fig. 4c) and iron (Fig. 4d) administration during adipocyte differentiation induced HMOX1 mRNA levels. Iron administration in pre-adipocytes (Fig. 4e) and fully differentiated adipocytes (Fig. 4f) also resulted in increased HMOX1 mRNA levels, whereas iron chelation using DFO (20 μmol/l) exerted opposite effects. As expected, the major inductor of HMOX1 mRNA levels was hemin (60 μmol/l), both in fully differentiated adipocytes (Fig. 4g) and during adipocyte differentiation (Fig. 5a). A significant reduction in intracellular lipid accumulation (Fig. 5b) and reduced adipogenic (ADIPOQ, FABP4, SLC2A4 and PLIN1) gene expression (Fig. 5c–f) was observed after hemin administration. Increased inflammatory (IL8, TNF) mRNAs (Fig. 5g–h), increased markers of intracellular iron accumulation (FTL, CYBA and SLC40A1) and reduced TFRC1 mRNA levels were also observed (Fig. 5i–l). Functionally, hemin led to a significant decrease in glucose uptake (Fig. 5m), basal respiration (Fig. 5n) and spare respiratory capacity (Fig. 5o).

Fig. 4
figure 4

HMOX1 gene expression in time course of adipocyte differentiation (a) and in adipose tissue subfractions (b). *p < 0.05 compared with adipocytes; **p < 0.01 compared with adipocytes or day 0. Effects of MCM (2.5%) (c), FeSO4 (Fe, 3 and 30 μg/ml) and DFO (20 μmol/l) (d) on HMOX1 gene expression during adipocyte differentiation at day 14. *p < 0.05, **p < 0.01 compared with differentiated adipocytes. Effects of FeSO4 (Fe, 3 and 30 μg/ml) and DFO (20 μmol/l) on HMOX1 gene expression in pre-adipocytes (e) and in adipocytes (f) after 72 h. Effects of hemin (60 μmol/l) on HMOX1 gene expression in adipocytes after 48 h (g). *p < 0.05, **p < 0.01 compared with vehicle. Diff, differentiated; non-diff, non-differentiated

Full size image
Fig. 5
figure 5

Effects of hemin (60 μmol/l) administration during adipocyte differentiation on HMOX1 gene expression (a), intracellular lipid accumulation (×10 magnification) (b), adipogenic-related (ADIPOQ, FABP4, SLC2A4, PLIN1) (c–f), inflammatory-related (IL8, TNF) (g–h), iron accumulation-related (FTL, CYBA, SLC40A1, TFRC) (i–l) gene expression markers, intracellular 2-DG6P in basal (white bars) and insulin stimulated (grey bars) conditions (m) and on mitochondrial respiratory function (basal respiration [n] and spare respiratory capacity [o]) at day 14. *p < 0.05, **p < 0.01 compared with differentiated adipocytes. † p < 0.05 compared with basal. Diff, differentiated; non-diff, non-differentiated; OCR, oxygen consumption rate

Full size image

Discussion

In contrast to haem oxygenase 2 (HMOX2), which is constitutively expressed in many tissues, HMOX1 is a facultative gene and is expressed at a relatively low level in most tissues. HMOX1 expression is regulated at transcriptional level being mainly induced by its substrate (haem), iron, oxidative stress and inflammation [26]. Facultative genes need precise regulation and their mRNA levels generally correlated with protein levels. In fact, a large number of studies have demonstrated that HMOX1 activity is proportional to HMOX1 mRNA and protein levels [27, 28]. Thus, we investigated HMOX1 mRNA levels as the main outcome in the current study; the association between HMOX1 mRNA and HMOX1 protein levels was also demonstrated in a subgroup of participants.

To the best of our knowledge this is the first study examining the possible role of HMOX1 in iron accumulation, mitochondrial respiratory function and adipogenesis in human adipocytes and adipose tissue in obese and non-obese participants.

HMOX1 and markers of iron accumulation in adipose tissue

Iron accumulation in adipose tissue in obesity is detrimental for adipogenesis [7,14,, 9, 10, 13–15]. In the current study, hemin-induced HMOX1 activity resulted in raised intracellular iron accumulation-related (FTL, SLC40A1 and CYBA) markers and concordantly decreased TFRC mRNA levels. In agreement with these data, the association between tissue iron and HMOX1 gene expression and activity has been demonstrated in brain [29, 30], aorta [31] and liver [32]. Furthermore, a positive association between total body iron and HMOX1 activity has been reported [33]. In addition, intracellular free-iron-produced reactive oxygen species (ROS) are important inductors of HMOX1 activity [27, 28]. For this reason, a bidirectional regulation between HMOX1 activity and intracellular iron levels should be considered. In fact, in this study (Fig. 4d–f) iron administration in pre-adipocytes and adipocytes and during adipocyte differentiation resulted in a significant upregulation of HMOX1 gene expression, which was attenuated by DFO (an iron chelator) administration. HMOX1 gene expression was positively correlated with gene expression markers of adipose tissue iron accumulation (FTL, SLC40A1 and CYBA [9]) in association with obesity in two cross-sectional independent cohorts. HMOX1 gene expression and HMOX1 protein were also associated with adipose tissue iron levels. The reduction of HMOX1 gene expression after bariatric surgery weight loss was positively correlated with the reduction of SLC40A1 and FTL. Considering that HMOX1 activity is an important source of intracellular iron, through the conversion of haem into ferrous iron [21], these findings suggest that adipose tissue HMOX1 is a marker of adipose tissue iron accumulation. In addition, the association between HMOX1 and markers of oxidative stress (CYBA, NFE2L2) in both SAT and VAT does not exclude a compensatory effect of HMOX1 through its anti-oxidative properties [26, 27].

HMOX1 and adipocyte mitochondrial function

Hemin-induced HMOX1 activity during adipocyte differentiation resulted in a significant alteration in mitochondrial respiratory function, decreasing basal respiration and spare respiratory capacity, two functional measures of mitochondrial function [21, 34]. Supporting these data, in this study HMOX1 gene expression was negatively correlated with gene expression markers of mitochondrial biogenesis (PPARGC1A, TFAM) even after controlling for age and BMI. Previous studies demonstrated the negative effects of HMOX1 on mitochondrial function in macrophages [21] and hepatocytes [35]. Since iron overload impairs mitochondrial function [36,37,38], it seems plausible that HMOX1 activity-associated adipocyte iron accumulation might contribute to attenuated mitochondrial respiratory function. The importance of mitochondrial function in adipocyte differentiation [39,40,41], adipocyte physiology [34], and the possible role of iron homeostasis in mitochondrial function and adipogenesis [10, 15] is well known.

HMOX1 and human adipogenesis

Hemin-induced HMOX1 expression resulted in decreased lipid droplet development in parallel with a significant reduction in adipogenic gene expression. A recent study also found anti-adipogenic effects of HMOX1 in pigs [42]. Specifically, this study demonstrated that overexpression of human HMOX1 in pigs resulted in decreased subcutaneous fat tissue development and attenuated capacity of isolated adipose-derived stem cells to differentiate to adipocytes [42].

Interestingly, HMOX1 induction led to a significant reduction of basal and insulin-induced glucose uptake in concordance with the decrease in SLC2A4 (GLUT4) mRNA levels. Since glucose uptake is required for the generation of new adipocytes and adipogenesis [43,44,45], the negative effects of hemin administration on glucose uptake might explain in part these anti-adipogenic effects. In agreement with these experiments, HMOX1 gene expression decreased significantly during human adipocyte differentiation and was negatively correlated with gene expression markers of adipogenesis (PPARG, FASN, ACACA) and adipocyte insulin action (SLC2A4, IRS1) in both VAT and SAT. Furthermore, in contrast with the expression of adipogenic genes, increased HMOX1 mRNA levels were found in parallel with hyperglycaemia and HOMA-IR in obese participants. Weight gain led to increased Hmox1 in mice [21] and rats (current results). Consistently, bariatric surgery-induced weight loss resulted in decreased adipose tissue HMOX1 mRNA. The change in HMOX1 mRNA levels correlated with changes in BMI, fat mass and adipose tissue markers of adipogenesis. However, no correlations were found between adipose tissue HMOX1 gene expression and systemic insulin sensitivity. As mentioned above, the association between HMOX1 expression and insulin resistance is not clear [21,22,23]. Biliverdin and carbon monoxide are two additional products derived from HMOX1 activity that have been associated with improved glucose tolerance and insulin sensitivity [46, 47]. Possibly, these putative positive effects of biliverdin and carbon monoxide might partly counteract the negative effects of HMOX1-associated adipose tissue dysfunction on systemic insulin sensitivity. The systemic upregulation of the haem oxygenase system by hemin or by cobalt protoporphyrin improves some metabolic features in Zucker diabetic fatty rats [48] and in obese Wistar rats [49]. Specifically, in Zucker diabetic fatty rats, hemin induction of the haem oxygenase system normalised glucose levels and reduced perirenal adiposity and inflammation [48]. However, these positive effects were attenuated in Zucker lean controls [48]. In obese Wistar rats, cobalt protoporphyrin-induced HMOX1 exerted positive effects on kidneys improving endothelial dysfunction in association with increased circulating adiponectin levels [49]. These studies investigated the systemic effects of HMOX1 upregulation in rats, whereas the current study focused specifically in adipose tissue HMOX1 expression in response to HFD-induced weight gain (in Wistar rats) or bariatric surgery-induced weight loss (in humans). For instance, current results demonstrated increased eWAT Hmox1 in association with total weight, circulating leptin and decreased Adipoq gene expression in rats. For this reason, our current findings were not comparable with those obtained in the earlier studies. Nevertheless, a possible protective role of adipose tissue HMOX1 in humans cannot be excluded. Further studies are required to investigate specifically the effects of adipose tissue HMOX1 induction on obesity-associated metabolic disturbances and insulin resistance.

HMOX1 and adipose tissue inflammation

As expected, we observed bidirectional influences between HMOX1 and inflammatory factors [21]. MCM (enriched with a mixture of inflammatory cytokines) induced adipocyte HMOX1 mRNA levels, while hemin-induced HMOX1 activity increased adipocyte inflammatory cytokines (TNF, IL8). HMOX1 mRNA levels were associated with gene expression of inflammatory cytokines in human adipose tissue and negatively correlated with CD206/CD68 ratio (an M2 macrophage marker). In rats, HFD-induced weight gain resulted in increased Hmox1 mRNA levels in association with Il6 in eWAT. Some products derived from the HMOX1 catalytic activity also promoted adipose tissue macrophage infiltration and inflammation in mice fed an HFD [50].

Current findings regarding the importance of haem in adipogenesis [20], the precise regulation in intracellular haem levels through HMOX1 activity required to avoid ROS-induced cellular damage [26, 27], and the importance of iron and its fine-tuned regulation in adipogenesis [7, 10] envision a possible scenario, in which a progressive increase of fat accretion (parallel to obesity) leads to increased intracellular haem levels and, in consequence, to induced HMOX1 gene expression and activity. Since HMOX1 is an inducible enzyme required for haem degradation, increased HMOX1 activity enhances the conversion of haem into ferrous iron [21], resulting in obesity-associated adipose tissue iron accumulation, which promotes adipose tissue dysfunction [7, 12].

In conclusion, HMOX1 is proposed as an important marker of iron excess-induced adipose tissue dysfunction in human obesity.