Circulating perilipin 2 levels are associated with fat mass, inflammatory and metabolic markers and are higher in women than men

cPLIN2 levels are higher in old women and positively associated with metabolic parameters and inflammatory markers

In order to study possible age-related modifications of cPLIN2 levels, we first analyzed a group of 189 subjects (Cohort 1). In particular, we compared a group of 35 young subjects (age range 24-39) with 154 old subjects, including centenarians, divided into four groups: 57 OLD1 subjects (age range 60-69); 33 OLD2 subjects (age range 70-79); 26 OLD3 subjects (age range 85-99); 38 centenarians (age range 100-107). The results of cPLIN2, as well as of anthropometric and biochemical parameters are reported in Table 1. cPLIN2 levels are significantly higher in women than men, and this difference becomes more dramatic in the group of old subjects (OLD1 and OLD2), however nonagenarians (OLD3) and centenarians show no difference.

Since BMI strongly correlates with cPLIN2, in both women (Spearman correlation parameters: ρ=0.572, p<0.001) and men (Spearman correlation parameters: ρ=0.712, p<0.001), further analyses were conducted considering BMI as a covariate. After adjusting for BMI, the difference in cPLIN2 levels by sex was confirmed, in particular women belonging to young, OLD1 and OLD2 groups showed significantly (p<0.001) higher levels of cPLIN2 with respect to men of the same age groups.

cPLIN2 levels appear to change with age, in particular they seem to be higher in women belonging to OLD1 and OLD2 groups with respect to young, nonagenarian, or centenarian subjects (Table 1), however, after adjusting for BMI, the difference between age groups disappeared. An apparent difference was noticed between young and OLD2 males, despite the similarity in BMI values. However, this difference resulted not statistically significant. A partial correlation of cPLIN2 with anthropometric, metabolic and inflammatory parameters after adjusting for BMI was then performed. The levels of cPLIN2 correlated positively with insulin and HOMA-IR, in both men and women (Table 2). Noteworthy, a strong correlation of cPLIN2 was also found with a marker of inflammation, such as leptin (Table 2). On the basis of the observed correlation with parameters related to diabetes, we checked whether people with a diagnosis of Type 2 Diabetes (T2D) have higher cPLIN2 levels than euglycemic ones. In OLD1, OLD2, OLD3 and Centenarians groups there were 36, 13, 2 and 3 subjects with T2D, respectively (Table 1). To our surprise, these subjects resulted to have cPLIN2 levels similar to those of the euglycemic subjects of the same age group (women No T2D: 47.04±27.5, T2D: 49.11±18.2; men No T2D: 26.35±21.4, T2D 31.81±17.3, cPLIN2 values are in ng/ml and expressed as mean±SD). The ANCOVA analysis considering BMI as covariate confirmed that no significant difference between the two groups is present. However, due to the small size of our study group, we cannot rule out that T2D can be a cause of increased secretion of cPLIN2.

Finally, in order to understand the possible interference of drug therapies on the levels of cPlin2, we compared the levels of cPLIN2 in subjects with or without lipid lowering, anti-diabetic or anti-hypertensive therapies. We analysed men and women belonging to OLD1 and OLD2 groups for which data on medications were available. As reported in Supplementary Table 1, the presence of anti-diabetic and anti-hypertensive therapies does not affect cPLIN2 levels. An apparent difference was noticed in patients undergo a lipid lowering therapy with respect to subjects that do not undergoing therapy, however after adjusting for BMI this difference resulted not statistically significant.

cPLIN2 levels are associated with parameters of body composition

As BMI is a synthetic index of body composition, we wondered whether cPLIN2 correlates with specific parameters of fat mass (FM) or lean mass (LM). In order to deepen this aspect, we analyzed cPLIN2 levels in a group of 129 additional and well-characterized old subjects (cohort 2): 63 men and 66 women of similar age (age range: 65-79 years) from the NU-AGE Project [17]. The subjects of this cohort were selected on the basis of their BMI (range 20-30) and total FM measured by whole-body DXA scan [18, 19]. These subjects were divided into two groups: low FM and high FM (see Table 3 for FM ranges). Most of the anthropometric and biochemical measures considered were significantly different between the two FM groups, in both men and women (Table 3).

The results of cPLIN2 in low and high FM groups are reported in Figure 1. cPLIN2 levels are significantly higher in women with respect to men, confirming the results showed in Table 1. Moreover, cPLIN2 is dramatically elevated in high FM group, in both men and women. Also in this case, since BMI strongly correlates with cPLIN2 in both women (Spearman correlation parameters: ρ =0.670, p<0.001) and men (Spearman correlation parameters: ρ =0.725, p<0.001), after adjusting for BMI, a partial correlation of cPLIN2 with anthropometric and metabolic parameters was performed (Table 4). Regarding biochemical parameters, a correlation between cPLIN2 and triglycerides is present only in women, while a weak correlation between cPLIN2 and glucose metabolism parameters (glycemia and HOMA-IR) was found only in men. This partial discrepancy with the data of Table 2 is likely due to the fact that Cohort 2 is more homogenous in terms of age range and health status [17]. Regarding body composition parameters, in both men and women positive and negative correlations were found with FM and LM parameters, respectively. Interestingly, hip circumference correlated positively in women and negatively in men, while the android/gynoid FM ratio was uncorrelated in women and positively correlated in men. A further focus on the fat of the android region indicated that the subcutaneous but not the visceral adipose tissue is strongly correlated with cPLIN2 in both sexes. Finally, a strong correlation of cPLIN2 with leptin and with alpha-1-acid glycoprotein (AGP) was found, this latter only in women (Table 4).

Subjects

Serum samples from a total of 318 volunteers of Caucasian ancestry, in the age range from 24 to 107 years were analyzed for this study.

A first analysis was performed on 189 subjects (cohort 1), divided into five age groups: young subjects (24-39 years), OLD1 subjects (60-69 years), OLD2 subjects (70-79 years), OLD3 subjects (85-99) and Centenarians (100-107 years) enrolled in Italy in the framework of the following projects: Italian National Project PRIN09 and EU Project GEHA. The study protocols were approved by the Ethical Committee of Sant’Orsola-Malpighi University Hospital, Bologna, Italy (Ethical clearance: 118/2004/U/Tess for GEHA Project; 22/2007/U/Tess for PRIN09 Projects). All subjects signed informed consent before entering the study. Exclusion criteria were the presence of malignant neoplasia and/or current therapy with immune suppressor drugs (like cyclosporine, methotrexate, and glucocorticoids) or anticoagulant drugs. A standard questionnaire was administered to the volunteers by nurse staff in order to collect lifestyle data, health status, clinical anamnesis, and details on medications. As for centenarians, in the case that the subject was unable to respond autonomously because of hearing or sight problems, the interview was performed with a relative or a caregiver. Young subjects were free of clinically evident diseases, while, as expected, old subjects and centenarians were reported to be affected by common age-related diseases, mostly hypertension, hypercholesterolemia and diabetes (Table 1).

The second step of analyses was conducted in additional 129 well-characterized old subjects (65-79 years), 66 women and 63 men (cohort 2), divided into two groups according to their total fat mass (FM), low and high FM, measured by Dual-energy X-ray absorptiometry (DXA) (Table 3). Also, in this case, the majority of the subjects were suffering from the most common age-related diseases (Table 3). These subjects are part of the Italian cohort of the EU project NU-AGE (registered with https://clinicaltrials.gov/, NCT01754012) and the recruitment of participants has been described in detail previously [17]. Briefly, volunteers aged 65-79 years from the community, free of major overt chronic diseases for at least 2 years (e.g., cancer, severe organ disease), living independently, and free of dementia, were recruited to participate in the baseline assessment. At enrollment, exclusion criteria included severe heart diseases, type 1 and insulin-treated type 2 diabetes, chronic use of corticosteroids, recent use of antibiotics, change in habitual medication use, frailty, malnutrition [body mass index (BMI) <18.5 kg/m2 or 10% weight loss within 6 months], or food allergy/intolerance requiring special diets.

Written informed consent was collected from all participants prior to their inclusion in the study, in accordance with the Declaration of Helsinki. NU-AGE was approved by the ethics committee of the S. Orsola-Malpighi Hospital Bologna, Italy (NU-AGE ethical approval n° 103/2011/U/Sper; 17/01/2012).

Sampling and data collection

Blood was drawn in the morning after an overnight fast. All samples were processed to collect serum. Serum was obtained after clotting and centrifugation at 760g for 20 minutes at 4° C, rapidly frozen and stored at −80° C.

Total and HDL cholesterol, triglycerides, C-reactive protein (CRP), glycaemia, insulin, were measured in serum by standard biochemical assays. Insulin resistance status was assessed using the homeostasis model assessment of insulin resistance (HOMA-IR), according to the previously described formula [29]: insulin (in microunits per milliliter) × glucose (in millimoles per liter)/22.5.

Body mass index (BMI) was calculated as weight in kilograms divided by the square of the height in meters (kg/m2). Waist (WC) and hip circumference (HC) were measured using a flexible steel tape as described by Bucci and coworkers [30]. Briefly, WC was measured at the end of exhalation, by wrapping the tape at the level of the iliac crest and umbilicus, with the subject standing. HC was measured at the level of maximal protrusion of the gluteal muscles. Handgrip strength test was performed to measure the maximum isometric strength of the hand and forearm muscles using a hand-held dynamometer (SMEDLYS’ dynamometer, Scandidact, Kvistgaard, Denmark) for two performances with each hand.

Subjects included in the cohort 2 of this study underwent a whole-body DXA scan to measure total and regional body composition using the following narrow-angle fan-beam densitometer: Lunar iDXA, GE Healthcare, Madison, WI, USA – enCORETM 2011 software version 13.6; BMD CV: ≤ 1.0% (Bologna, Italy) as previously reported [18, 19]. Briefly, the region of interests included six different corporeal regions: total body, trunk, upper limbs, lower limbs, android region and gynoid region. For each region, DXA scanned the weight (in g) of total fat, lean and bone mass: whole body FM, non-bone whole body lean mass (LM), bone mineral content (BMC) and density (BMD). Measurements of visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) were obtained at android level with CoreScan software. The relationship between parameters was investigated using specific reliable indexes: total body FM/LM, Fat Mass index (FMI, whole body fat mass/height2), Lean Mass index (LMI, whole body lean mass/height2), android/gynoid FM, android FM/LM, Appendicular Lean Mass index (ALMI, lean mass from arms plus legs/height2) and Skeletal Mass Index (SMI, lean mass from arms plus legs/weight) [31].

Markers of inflammation, adipose related hormones and cPLIN2

Inflammatory and adipose-related markers were analyzed by a magnetic bead-based multiplex immunoassays (Bio-Plex) (Bio-Rad) according to the manufacturer’s instructions. In particular Ghrelin (inter-assay CV, 2%) and Resistin (inter-assay CV, 4%) in multiplex with Bio-Plex Pro human diabetes assay. Plates were read and analyzed by Bio-Plex Manager Software.

The quantitative determination of high sensitivity C-reactive protein (hsCRP), leptin, adiponectin has been performed by ProcartaPlexTM Immunoassay (Thermo Fisher Scientific) according to the manufacturer’s instructions. Analysis was performed using Luminex 200 instrumentation (Luminex Corporation). Assay sensitivities were 19.31 pg/mL for Leptin, 4.39 pg/mL for hsCRP, and 47.46 pg/mL for adiponectin. α1 Acid glycoprotein (AGP) has been measured by an immunoturbidimetric assay (AAGP2, Tina-quant α1-Acid Glycoprotein Gen.2 COBAS, Roche Diagnostics) with a measuring range of 0.1-4.0 g/L.

cPLIN2 concentration was determined in serum samples by ELISA assay using the commercial Human ADRP ELISA Kit (E-EL-H0278), according to the manufacturer’s instructions. After preliminary testing, serum was preferred to plasma, and samples were diluted 1:8 with the sample diluent provided by ELISA kit as this dilution resulted in the one yielding the most stable results. The intra- and inter-assay CV range were: 4.6% – 0.1% and 5.2% – 3.3%, respectively. In all the samples, PLIN2 was measured in duplicate and the mean values were used in the statistical analyses. The standard curve was determined by simultaneously analyzing a dilution series of standard samples. The final data were obtained in a blind set up by the operator. Synergy™ fluorometer (Bio-Tek Instruments, Winooski, Vermont, USA) was used to read the absorbance of each plate.

Statistical analysis

The data were analysed with non-parametric tests since they did not follow a normal distribution. In particular, the comparison among the age groups in the cohort 1 was performed by using Kruskal-Wallis test, while Mann-Whitney test was used in cohort 1 for the comparisons between men and women, and in cohort 2 for the comparisons between low FM and high FM groups, as well as between men and women of the same FM groups. The Bonferroni correction was applied. Analysis of covariance (ANCOVA) was used to compare the mean differences in cPLIN2 levels after adjustment for BMI. The relationships between cPlin2 levels and anthropometric and biochemical parameters were calculated by Spearman rank correlation test and partial correlation analysis after adjusting for BMI. Significance was accepted as p < 0.05. Data are expressed as mean ± SE or SD. All data were analysed using the SPSS 23.0 for Windows software (SPSS Inc.; Chicago, IL, USA).