Obesity is an important global health problem. According to the World Health Organization (WHO), obesity prevalence rates tripled between 1975 and 2016, and currently, around 650 million people worldwide (15% of women and 11% of men aged ≥18 years) suffer from excessive body weight [1
]. Obesity is not only a cosmetic defect but can also cause a variety of metabolic abnormalities, leading to the deterioration of health and quality of life. It is well known that excessive body weight is associated with insulin resistance and hyperinsulinemia. Consequently, obesity is recognised as an important risk factor in the pathogenesis of type 2 diabetes mellitus [2
]. Furthermore, obesity also increases the risk of developing cardiovascular diseases, hypertension, dyslipidaemia and several other abnormalities associated with the development of metabolic syndrome [3
Metabolic syndrome represents several cardiovascular risk factors associated with obesity, such as disturbed glucose and insulin homeostasis, atherogenic dyslipidaemia and arterial hypertension. These components of metabolic syndrome are also complemented with chronic low-grade inflammation, coagulopathy, endothelial dysfunction and oxidative stress [4
]. The prevalence of metabolically unhealthy obesity is estimated to be around 75%, while approximately 25% of obese subjects are metabolically healthy. Unfortunately, factors associated with healthy and unhealthy obesity still remain unclear. Nevertheless, early identification of subjects at high risk of developing metabolic abnormalities might allow appropriate preventive and curative measures to be taken [5
The metabolic syndrome usually develops as a consequence of increased energy intake and low level of physical activity, with other important factors including diet composition, aging population and genetic background [6
]. Previously, several genes have been identified as potentially associated with different features of the metabolic syndrome, such as transcription factor 7-like 2 (TCFL2
), fat mass and obesity gene (FTO
), adenylate cyclase type 5 (ADCY5
), fatty acid desaturase 1 (FADS1
), GLI-similar 3 (GLIS3
), insulin-like growth factor 1 (IGF1
) and peroxisome proliferator-activated receptor-gamma (PPARγ
It has been hypothesised that variants in lactoferrin (LTF
) and lactoferrin receptor-related genes may also play an important role in the development of metabolic abnormalities in obese subjects. Previously, LTF
gene polymorphisms were reported to be associated with lactoferrin levels and coronary artery stenosis [10
]. Moreover, LTF
rs1126477 and rs1126478 polymorphisms were reported to be associated with high-density lipoprotein cholesterol (HDL-C) and triglycerides (TG) levels in subjects with impaired glucose tolerance [11
]. Interestingly, in children, three single nucleotide polymorphisms (SNPs)—rs1126478, rs34827868 and rs1042073—in the LTF
gene had a minor allele associated with increased HDL-C concentrations and three others (rs4637321, rs2239692 and rs10865941) were related with decreased fasting glucose levels, increased blood pressure, and higher levels of free fatty acids. However, these associations did not remain significant after correction for multiple testing [12
]. Another study demonstrated that the hypolipidemic effect of lactoferrin is dependent on the selective binding of this protein to the low-density lipoprotein receptor-related protein 1 (LRP1). It was assumed that modifications of the lactoferrin molecule affect the interaction with LRP1 receptors, thus influencing the level of lipids and the rate of their removal from the circulation. Indeed, the Arg-rich sequence of the N-terminus of the lactoferrin resembles the structure of apolipoprotein E recognised by LRP1 [13
]. In addition, some gene variants in LRP1
rs4759277 are associated with insulin concentrations and homeostasis model assessment of insulin resistance (HOMA-IR) [7
]. However, whether the above polymorphisms in lactoferrin-related genes influence the development of metabolic syndrome in obese subjects remains unclear.
Therefore, the study aimed to assess the prevalence of selected LTF and lactoferrin receptor gene polymorphisms in metabolically healthy obese (MHO) and metabolically unhealthy obese (MUHO) subjects, examining the impact of analysed gene polymorphisms on individual components of the metabolic syndrome.
2. Materials and Methods
2.1. Study Population
Obese men and women were recruited to the study at the Poznan University of Medical Sciences, Poland. During the admission process, subjects received information about the study, its aim, possible benefits and risks. Study participants were informed that participation in the study was voluntary and that they may refuse to participate or discontinuing participation at any time without giving reasons. Written informed consent was obtained from all participants.
In total, 452 obese subjects were included in the study and were divided into two groups: MHO and MUHO group. The International Diabetes Foundation criteria were used to identify metabolic syndrome [14
]. The primary inclusion criteria were as follows: body mass index (BMI) ≥ 30 kg/m2
or waist circumference (WC) ≥ 80 cm for women, and ≥94 cm for men. In addition, during dividing the study population into two groups, the following inclusion criteria were used: (1) MHO: less than two of the following disorders, (2) MUHO: at least two of the following disorders:
systolic blood pressure (SBP) ≥ 130 mmHg and/or diastolic blood pressure (DBP) ≥ 85 mmHg and/or antihypertensive therapy;
TG levels ≥ 150 mg/dL (1.7 mmol/L) and/or specific treatment for this lipid abnormality;
fasting glucose levels ≥ 100 mg/L (5.6 mmol/L) and/or previously diagnosed type 2 diabetes;
HDL-C levels < 40 mg/L (1.03 mmol/L) in men, <50 mg/dL (1.29 mmol/L) in women and/or specific treatment for this lipid abnormality [4
The exclusion criteria were as follows:
cancer diagnosis in the last 5 years;
general poor health status;
pregnant and breastfeeding women.
This study was conducted according to the Declaration of Helsinki. The study protocol was approved by the Poznan University of Medical Sciences Bioethical Committee (refs. 984/17, 1161/19).
2.2. Assessment of Anthropometric Parameters
The following anthropometric parameters were assessed in this study: body weight, body height and WC. During anthropometric measurements, all participants wore light clothes and were barefoot. Body height was measured in the standing position and rounded up to the nearest 0.5 cm. Body weight and body height were measured by a calibrated electronic scale with a stadiometer. WC was assessed on the bare skin between the lowest rib and the iliac crest, during minimal respiration. In this study, the WHO criteria were used to defined abdominal obesity (WC ≥ 94 cm in men and ≥ 80 cm in women) [15
], measured using a standard tape measure. BMI was calculated based on body weight and body height, defined as body weight in kilograms divided by body height in meters squared and classified according to the WHO criteria [16
2.3. Blood Pressure
Blood pressure was measured according to the European Society of Hypertension guidelines. SBP and DBP were measured on the arm at heart level and were expressed by three measurements. Normal blood pressure is defined as SBP < 130 mmHg and DBP < 85 mmHg, while hypertension is defined as SBP ≥ 140 mmHg or DBP ≥ 90 mmHg for most adults [17
2.4. Blood Collection and Biochemical Measurements
Blood samples were taken fasting by registered staff nurses. All biochemical parameters were measured using standardised laboratory methods, including fasting plasma glucose concentrations, serum concentrations of insulin, total cholesterol (TC), HDL-C, low-density lipoprotein cholesterol (LDL-C), TG and C-reactive protein (CRP; data available for 450 subjects; for 258 subjects exact CRP levels were available, for 192 subjects CRP levels were measured by standard methods and the analyser was not able to detect values < 4 mg/L).
Genomic DNA was isolated from EDTA anticoagulated blood according to the membrane-based DNA extraction protocol (Blood Mini, A&A Biotechnology, Poland, or equivalent). In short, defrosted blood samples of 1 mL were mixed with 500 μL LE and then centrifugated at 10,000–15,000 revolutions per minute (rpm) for 3 min. Then, the supernatant was discarded, 100 µL of Tris buffer was added and cells were resuspended by pipetting. For the prepared samples 20 μL protease K and 200 μL buffer LT were added. After incubation at 37 °C for 20 min, 20 s of the vortex, 1 min of centrifugation at 10,000–15,000 rpm and discarded the filtrates, 500 μL of A1 wash solution was added. The mixture was centrifuged at 10,000–15,000 rpm for 1 min, and another 400 μL of A1 wash solution was added. After centrifuging at 10,000–15,000 rpm for 2 min, 100 μL of Tris elution buffer heated to 75 °C was added. The mixture was incubated for 5 min at room temperature and centrifuged for one minute at 10,000–15,000 rpm. The yielded DNA concentrations were measured using a NanoDrop™ One Spectrophotometer (ThermoScientific, Wilmington, NC, USA). SNPs in LTF rs1126477, LTF rs2239692, LTF rs1126478, low-density lipoprotein receptor-related protein 2 (LRP2) rs2544390, LRP1 rs4759277, and LRP1 rs1799986, genes were genotyped with TaqMan allelic discrimination assays (Applied Biosystem, Foster City, NC, USA): rs2544390: C___8822318_10, rs4759277: C___31186847_10, rs1799986: C___1955081_10, rs1126477: C___9698511_10; rs2239692: C___2610649_10; rs1126478: C___9698521_10) using Bio-Rad CFX96™ Real-Time PCR system (Hercules, CA, USA). The PCR reactions (10 μL) contained: (1) TaqPath ProAmp Master Mix—5 μL, (2) TaqMan Genotyping Assay (containing sequence-specific forward and reverse primers) —0.5 μL, (3) DNA—4.5 μL, (4) Nuclease free water—10 μL. The 96-well plates were used. In 95 wells, genomic DNA was disposed of. Each plate consisted of one negative control. After the addition of the TaqPath ProAmp Master Mix, the plate was covered with the PCR plate sealer and briefly centrifuged in the plate centrifuge. Amplification conditions were as follows: (1) Pre-read run—30 s—60 °C—hold, (2) Enzyme activation—5 min—95 °C—hold, (3) Denaturation—15 s—95 °C—40 cycles, (4) Anneal/extend—1 min—60 °C—40 cycles, (5) Post-read run—30 s—60 °C—hold.
2.6. Statistical Analysis
The STATISTICA 12.0 PL (StatSoft, Tulsa, OK, USA) and PQStat (PQStat Software, Poznań, Poland) software (α = 0.05) were used for the statistical analysis. A two-sided p
-value < 0.05 was considered as statistically significant. The overall characteristics of subjects were expressed as medians and interquartile ranges (IQRs), means and standard deviations (SDs) or as frequencies and percentages. The normality of the distribution of the variables was verified using the Shapiro–Wilk test. Allele and genotype frequencies of the analysed polymorphisms were tested for consistency with Hardy-Weinberg equilibrium using exact tests. De Finetti diagrams with Hardy-Weinberg parabola were generated using the online programme tool [18
]. Allele frequency differences were assessed by the Chi2
test and genotype differences by Armitage’s trend test [18
]. Quantitative phenotypic traits were determined using Mann–Whitney U and Kruskal–Wallis tests. Post-hoc analysis was performed for pairwise comparisons of subgroups using Dunn’s test with Bonferroni correction. Contingency tables were used to assess relationships between categorical variables.
One of the key findings of this research were differences in genotype frequencies of the LTF rs2239692 between MHO and MUHO subjects. Moreover, we found that the CT variant compared to the TT variant of this polymorphism was associated with lower odds of developing metabolic syndrome. Furthermore, we demonstrated several associations between analysed gene polymorphisms and individual components of the metabolic syndrome. To the best of our knowledge, no studies have yet compared the prevalence of selected LTF and lactoferrin receptor genes polymorphisms in MHO and MUHO subjects.
In 1980, Andres first suggested that obese subjects should be classified into two groups: MHO and MUHO [19
], with the MHO group presenting a beneficial metabolic profile compared to the MUHO group. MHO is characterised by lower blood pressure, glucose levels and lipid profiles, as well as higher insulin sensitivity, compared to MUHO [20
]. In addition, MHO subjects have lower all-cause and cardiovascular disease mortality than MUHO subjects [21
]. Therefore, it is important to discriminate the two phenotypes of obesity. However, currently, there is a lack of consensus regarding defining MHO and MUHO subjects [22
]. Moreover, factors associated with healthy and unhealthy obesity phenotypes remain unclear [5
]. Here, we hypothesis that in addition to lifestyle factors, genetic factors might partly explain the differences between MHO and MUHO subjects. Previously, several studies identified genes which might be potentially associated with different features of the metabolic profile [7
]. Additionally, previous studies suggested that selected polymorphisms in LTF
, and LRP2
genes might be associated with the prevalence of metabolic abnormalities [7
gene is organised into 17 exons, ranging in size from 23 to 35 kb [27
] and is located on human chromosome 3, position 3p2112 [28
]. This gene is highly polymorphic with the presence of several common alleles in the general population [27
]. Previously, several studies have suggested that some SNPs in LTF
gene might be associated with the prevalence of metabolic abnormalities [11
]; however, none of the studies evaluated the association between LTF, LRP1
genes polymorphisms with the prevalence of metabolically healthy or metabolically unhealthy obesity. Similarly to our results, Marcil et al. [12
] in a study conducted on 1749 French Canadians aged 9, 13 and 16 years and found a significant difference in allele frequencies between subjects with and without metabolic syndrome for the LTF
rs2239692 polymorphism. However, the association did not remain significant after correction for multiple testing.
Our study is the first that demonstrated a significant association between the LTF
rs1126477 gene variants and the anthropometric parameters. More specifically, we noticed that the CT variant of the LTF
rs1126477 was associated with lower WC in the total population and lower BMI in the MHO group compared to the TT variant. However, our results contrast with those of a previous study conducted on the male population with normal blood glucose levels or an altered glucose tolerance and reported no association between the LTF
rs1126477 and rs1126478 polymorphisms and the anthropometric parameters. Interestingly, the researchers observed that subjects with a normal glucose tolerance who were AG heterozygotes for LTF
rs1126477 had significantly decreased TG levels. Similarly, G carriers for LTF
rs1126478 had significantly lower TG levels and significantly higher HDL-C levels than AA homozygotes. These associations remained significant after controlling for age, BMI, waist-to-hip ratio, fasting glucose concentrations, smoking status, and alcohol intake. In addition, the authors suggested that carriers of the G allele of LTF
rs1126478 may have a better ability to inhibit modified lipoprotein uptake in macrophages than carriers of the A allele [11
]. However, these findings were not confirmed in the present study. Nevertheless, in MHO subjects we found significant differences in LDL-C levels between LTF
rs1126477 gene variants.
Recently, the association between LTF
gene polymorphisms and blood pressure was reported, with Alexander et al. [23
] observing that LTF
rs1126478 was over-represented in subjects with hypertension compared to controls. Using a recessive genetic model, researchers found that the frequency of homozygosity for the minor allele (GG) in hypertensive group significantly increased relative to controls. In addition, for an additive genetic model, but not for dominant genetic model, researchers observed a trend for a significant association of LTF
rs1126478 with hypertension. In our study, for the first time we compared the effect of LTF, LRP1
genes polymorphisms on blood pressure in MHO and MUHO groups. We did not find differences between gene variants of LTF
rs1126478 polymorphism and blood pressure, but we showed significant differences in DBP between LRP1
rs1799986 gene variants in MUHO subjects.
Regarding putative lactoferrin receptors, LRP1 is an endocytic and signalling receptor which is widely expressed in several tissues. LRP1 is a member of the LDL receptor family which is involved in the clearance of chylomicron remnants from the circulation and present cardioprotective effect. The previous study demonstrated that LRP1 is involved in insulin and glucose homeostasis [29
]. Therefore, it was hypothesised that SNPs in the LRP1
gene might also affect the prevalence of metabolic abnormalities in obese subjects. Indeed, Delgado-Lista et al. [7
] evaluated the association of 904 SNPs selected for their potential contribution to carbohydrate metabolism in 450 participants in the LIPGENE cohort and found that fasting insulin, and C-peptide levels, as well as HOMA-IR, and the quantitative insulin sensitivity check index (QUICKI) significantly differed according to LRP1
rs4759277 gene variants. These results were in contrast to our findings, as we did not observe an association between LRP1
rs4759277 gene variants and glucose and insulin homeostasis. However, we noted significant differences in HDL-C concentrations between genetic variants of this polymorphism in the MHO group but not in the MUHO group. Besides, we noted that LRP1
rs4759277 polymorphism is associated with DBP in MUHO and HDL-C levels in MHO subjects. Previously, no studies have reported an association between LRP1
rs4759277 gene variants and lipid profile as well as blood pressure. Nevertheless, Aledo et al. [24
] found that LRP1
rs1799986 polymorphism in the dominant model (CT + TT vs. CC) was significantly associated with premature cardiovascular disease in familial hypercholesterolemia after adjusting for sex, age and BMI. Besides, Pocathikorn et al. [30
] found a significantly lower frequency of TT variant of LRP1
rs1799986 polymorphism in subjects with coronary heart disease compared to controls. In contrast, Benes et al. [31
] showed that subjects with the 5G/5G plasminogen activator inhibitor-1 genotype and the T allele had increased risk of coronary heart disease.
Recent studies have also suggested that LRP1
is a likely contributor to adipogenesis and adipocyte homeostasis. In addition, it has been shown that the expression of this gene in obese subjects in adipocytes is increased [32
]. Moreover, Hoffman et al. [35
] reported that LRP1
knockout mice have a lower fat mass and elevated energy expenditure, whereas Liu et al. [36
] showed that LRP1
knockout mice have a two-fold increase in fat mass compared to wild-type mice, which was associated with increased food intake, reduced energy expenditure and decreased leptin concentrations. The association of LRP1
knockout mice with increased fat mass was also supported by Terrand et al. [37
], who found that LRP1
knockout mice had a higher body fat which was associated with reducing lipolysis. Based on these results, we hypothesised that selected SNPs in the LRP1
gene might be associated with anthropometric parameters. Indeed, Frazier-Wood et al. [38
] observed that homozygous subjects for the minor allele at the LRP1
rs715948 polymorphism had BMIs around 1.03 kg/m2
higher than major allele carriers. In our study, we found an association between WC and LRP1
rs4759277 polymorphism in MHO subjects.
LRP2 encodes low-density lipoprotein receptor-related protein 2 (megalin) and is a member of the low-density lipoprotein receptor family. The receptor is expressed in the epithelial of renal proximal tubules, the epididymis, and thyroid cells and probably play an important role in the reabsorption of proteins and endocytosis [39
]. Previously, it was reported that the T allele of LRP2
rs2544390 polymorphism is associated with higher serum uric acid levels [40
]. In addition, Nakatochi et al. [25
] found that the T allele in LRP2
rs2544390 polymorphism was significantly associated with a higher risk of metabolic syndrome development in Japanese male employees, whereas Sun et al. [26
] noted that the T allele in LRP2
rs2544390 was significantly correlated with increased fasting insulin concentrations, HOMA-IR and the second-phase Stumvoll index. However, this study was conducted in Chinese women and it cannot be ruled out that in European subjects, other genetic factors affect glucose and insulin homeostasis. Indeed, in our study, we did not observe an association between LRP2
rs2544390 polymorphism and metabolic abnormalities.
CRP is a biomarker of inflammation which may constitute an independent risk factor for cardiovascular disease [42
]. In our study, we assessed the impact of analysed genes polymorphisms on CRP levels. However, we did not demonstrate any association. Nevertheless, we showed that in the MHO group statistically significant more subjects had CRP values < 4 mg/L than in the MUHO group (p
= 0.0120). However, no possibility to obtain high sensitivity CRP values for each included subjects could affect our findings. Our study has some limitations. The small number of tagging SNPs genotyped and the low prevalence of some of the analysed SNPs in the European population could result in limited power to detect significant gene-related associations. Furthermore, it is also possible to obtain false-positive results when several SNPs are analysed. Moreover, our results might be related to other SNPs in linkage disequilibrium with analysed polymorphisms. Furthermore, women were the majority of our population and our analysis was limited to white European descent. Therefore, it is not clear if these results are generalisable to other ethnicities. In addition, a relatively small number of the studied subjects received antihypertensive or hypolipemic drugs and none of the subjects received hypoglycemic treatment. However, the use of antihypertensive, hypolipemic and hypoglycemic therapies is usually common in the obese population. Besides, the MHO group was significantly younger than the MUHO group, so it is also probable that some subjects from the MHO group will develop metabolic abnormalities within a few years. Furthermore, we did not adjust for other confounding factors, such as diet, alcohol consumption and cigarette smoking, as these may bias the association between analysed SNPs and the prevalence of metabolic syndrome. Finally, we did not assess the effect of analysed SNPs on lactoferrin levels.
The strengths of this study are the well-characterised study population and the inclusion of many reliable biochemical parameters. For genotyping, we used TaqMan allelic discrimination assays, the simplest SNPs genotyping technology, which is easy to automate and scale up. Besides, this is the first study that compared the prevalence of selected LTF and lactoferrin receptor genes polymorphisms in MHO and MUHO subjects.