Subjects were recruited at Don Gnocchi Foundation, Florence, Italy. A written informed consent was obtained from each participant. The study was approved by the Ethics Committee of Don Gnocchi Foundation and performed following the Ethical standards of the Declaration of Helsinki, 1964 and its later amendments (Carlson et al., 2004). The number of ClinicalTrials.gov Identifier is NCT03441802.

Inclusion criteria: BMI ≥18.5 kg/m².

Exclusion criteria: Subjects suffering of eating disorders and in recent (1 month) or ongoing antibiotic therapy.

We enrolled 36 individuals, 17 males and 19 females, eligible in the study.

Subjects were divided in two arms according to their BMI: controls had normal weight (BMI between 18.5 and 24.9 kg/m²). They were 6 males and 12 females and at enrollment they were following MD, evaluated by a specific score (Sofi et al., 2014). Cases were overweight/obese subjects (BMI ≥25 kg/m²). Not all cases, only 11 males and 7 females, at enrollment, followed a MD according to the score criteria.

At enrollment, all the participants to the study received the same procedures and evaluations. At time 0 (T0) all the participants were submitted to anamnestic, anthropometric and dietetic assessments: height was measured on a wall-mounted stadiometer, weight was measured on a leveled platform scale and BMI was calculated as weight (kg)/height (m²). The adherence to the MD was evaluated by a specific score from 0 to 18 (0 minimal adherence; 18 greatest adherence) (Sofi et al., 2014). The diet was analyzed using the WinFood database (Medimatica, Teramo, Italy) according to the table of food consumption of the Italian National Institute of Nutrition and to the Food Composition Database for Epidemiologic Study in Italy. The survey items reported the quantities and frequency of food consumption such as cereals, vegetables and fruit, beans, fish, poultry, dairy products, milk, fats and oils, eggs, meat, sweets, salt, alcohol and other beverages (Salvini et al., 1996).

After enrollment all subjects followed a typical MD (55–60% carbohydrates, mainly complex ones, 25–30% polyunsaturated and monounsaturated fats, 15–20% proteins) (Luisi et al., 2015) and cases received a low-calorie MD (Kcal 1,552 ± 160). Both cases and controls utilized 40 g/die of HQ-EVOO for 3 months as the only cooking and dressing fat. HQ-EVOO main constituents are reported in Table 1 and the analysis were performed by the Valoritalia Laboratories s.r.l., Tavarnelle Val di Pesa, Florence, Italy.

Anthropometric assessments, questionnaires, blood and fecal samples were obtained at T0 and repeated after 3 months of MD rich in HQ-EVOO (T1).

The enrolled individuals followed these drug therapies: one case, affected by type 2 diabetes, received metformin (250 mg × 2/die); four participants had hypertension and were treated with ramipril (5 mg/die).

Ematochemical parameters: All blood samples were immediately placed on ice and centrifuged at 4°C, after which the plasma was removed. Plasma was stored at −80°C until analysis. Plasma glucose was estimated using a Beckton Dickinson blood sugar assay kit. Ten microliters of plasma was mixed with 90 µl of reagent consisting of glucose oxidase-peroxidase-O-dianisidine. After 10 min, the Optical density (OD) of the complex was measured at 620 nm in an autoanalyzer. Plasma insulin was measured using a human insulin ELISA kit (Novus Biologicals–Bio-Techne Ltd, Abingdon, UK); while plasma C-peptide concentrations were assayed by an ultrasensitive C-peptide ELISA kit (Mercordia AB8, Uppsala, Sweden). The assay was calibrated against International Reference Reagent for C-peptide (IRR-C-peptide, a WHO standard).

Determination of adiponectin levels: Adiponectin, an adipocyte-specific protein, which plays a role in the development of insulin resistance, was measured in plasma using a commercially available ELISA kit (Adipo Bioscience, Santa Clara, CA, USA). The assay was carried out according to the manufacturer procedures. The developed color was measured spectrophotometrically using the micro plate reader at 450 nm. Adiponectin concentrations, in µg/ml, were calculated from the standard curve prepared using recombinant human adiponectin standards.

Determination of 8-hydroxy-2-deoxy-guanosine (8-OHdG): levels of 8-OHdG, a marker of oxidative DNA damage, were measured on plasma (Lucarini et al., 2017). Briefly, frozen plasma samples were added with 1 ml of 10 mmol/l Tris–HCl buffer, pH 8, containing 10 mmol/l EDTA, 10 mmol/l NaCl, and 0.5% SDS, incubated for 1 h at 37°C with 20 µg/ml RNase 1 (Sigma‐Aldrich, Saint Louis, MO, USA) and overnight at 37°C under argon in the presence of 100 µg/ml proteinase K (Sigma‐Aldrich). The mixture was extracted with chloroform/isoamyl alcohol (10/2 v/v). DNA was precipitated from the aqueous phase with 0.2 volumes of 10 mmol/l ammonium acetate, solubilized in 200 µl acetate buffer, pH 5.3 and denatured at 90°C for 3 min. The extract was then supplemented with 10 IU of P1 nuclease (Sigma‐Aldrich) in 10 µl and incubated for 1 h at 37°C with 5 IU of alkaline phosphatase (Sigma‐Aldrich) in 0.4 mol/l phosphate buffer, pH 8.8. All of the procedures were performed in the dark under argon. The mixture was filtered by an Amicon Micropure‐EZ filter (Merck‐Millipore, Milan, Italy) and 50 µl of each sample was used for 8‐OHdG determination using a ELISA kit (JalCA, Shizuoka, Japan), following the instructions provided by the manufacturer. The absorbance of the chromogenic product was measured at 450 nm and expressed as ng/mg of DNA. The results were calculated from a standard curve based of 8‐OHdG solution. The values are expressed as ng 8‐OHdG/ng total DNA.

Determination of malonyldialdehyde (MDA): the measurement of lipid peroxidation in plasma samples was based on the reaction of MDA, the end product of the process, with 2-thiobarbituric acid (TBA) to form a chromophore absorbing at 532 nm. The reaction mixture (0.4 ml of sample; 0.2 ml of 2 mM chlortetramethoxypropane; 0.2 ml of 8.1% sodium dodecylsulphate; 1.5 ml of 20% acetic acid and 1.5 ml of aqueous solution of TBA up to 4 ml with distilled water) was heated to 95°C for 30 min and then 1 ml of N-butanol and pyridine (15:1 v/v) was added to plasma samples. The absorbance of the organic phase was measured at 532 nm (Mullane et al., 1985). The values were expressed as nanomoles of TBA-reactive substances (MDA equivalent) mg⁻¹ of protein, determined with the Bradford method (Bradford, 1976) over an albumin standard curve.

Determination of myeloperoxidase (MPO) activity: myeloperoxidase, a reliable marker of leukocyte activation, was determined as described in the literature (Masini et al., 2002) with some modifications: 100 ml of plasma was allowed to react with a solution 1.6 mM of tetramethylbenzidine and 0.1 mM H₂O₂. The rate of change in absorbance was measured spectrophotometrically at 650 nm. MPO activity was defined as the quantity of enzyme degrading 1 µmol of peroxide per minute at 37°C and was expressed in mU per mg of protein, determined with the Bradford method (Bradford, 1976) over an albumin standard curve.

Determination of the cytokines: the levels of two pro‐inflammatory cytokines, interleukin-6 (IL‐6) and tumor necrosis factor-α (TNF-α) and anti-inflammatory interleukin-10 (IL-10) were measured on aliquots (50 µl) of plasma by using the Flow Cytomix assay (Bender Medsystems GmbH, Vienna, Austria), following the protocol provided by the manufacturer. Fluorescence was read with a cytofluorimeter (CyFlow^® Space, Partec, Germany). Values are expressed as pg/µg of total proteins determined over an albumin standard curve (Bradford, 1976).

Total DNA (Agnelli et al., 2004) was extracted from fecal samples by following the QIAamp DNA Stool Mini Kit instructions (Qiagen) and quantified with a Qubit^® 2.0 fluorometer (Invitrogen, USA). Molecular weight and fragment length of DNA were checked on 1.5% agarose gel; the yield was calculated as µg DNAg⁻¹ feces. Quantitative PCR (qPCR) was conducted using the specific primers rpoB1, rpoB1o and rpoB2 for the RNA-polymerase β subunit encoding gene, a valuable alternative as present in single copy per genome (Dahllöf et al., 2000; Adékambi et al., 2008; Větrovský and Baldrian, 2013), that generate amplicons of 250-bp (Renouf et al., 2006), on 10 ng DNA for all the samples; negative control contained only ddH₂O. Reactions were performed in a CFX Connect 96 apparatus (BioRad, Hercules, CA, USA) and the results were analyzed by the manufacturer’s software. Amplification was carried out in a 25 µl final volume containing: 7.5 µM of each primer, 2.5 µg of BSA, 1X iTaq^™ Universal SYBR^® Green Supermix, sterile ddH₂O to reach the appropriate volume. Amplification was performed in 96-well micro titer plates (BioRad). The program cycle was: 95°C 3 min followed by 35 cycles of 95°C 1 min, 45°C 1 min and 72°C 1 min. After that, a melting curve program was run for which measurements were made at 0.5°C temperature increments every 10 s within a range of 60–100°C. A rpoB amplified and purified fragment (from a mix of LABs of a commercial probiotic formulation—Lactoflorene^® Plus, Montefarmaco OTC, (Soldi et al., 2019)) was used as standard.

The standard curve was developed by plotting the logarithm of known concentrations (tenfold dilution series in triplicate from 6.5 × 10^-1 to 6.5 × 10⁻⁵ in 25 µl reaction) of the rpoB fragment against the threshold Cycle (Ct) values (Ceccherini et al., 2003). The qPCR standard curve had an R² of 0.99–0.97 and an efficiency >85%. Three replicates were carried out for each sample for the two interval times: T0 and T1.

Sample size calculation was performed to assess the minimum number of participants to be included in the study (website: http://hedwig.mgh.harvard.edu/sample_size/size.html) (Schoenfeld, 2001). It was calculated considering that 80% was the probability that the study revealed a difference of MPO levels between the two groups (cases and controls at T0) with a significance level of 0.05, assuming that the set power was 0.8. The minimum sample size for this study was 18 subjects.

Data were reported as mean values of individual average measures of each subject. Delta (Δ) is the variation between T0 and T1. Delta values were calculated by subtracting the final value at 3 months of MD from the corresponding initial value at baseline. Significance of differences among the groups was assessed by t-test or repeated measures two-way analysis of variance followed by Bonferroni’s post-hoc test using GraphPad Prism 4.03 statistical software, when appropriated, and QuickCalcs (GraphPad Software, Inc, La Jolla, CA).

Body weight and BMI were significantly different at the same time point between controls and cases (controls T0 vs cases T0 and controls T1 vs cases T1). Moreover, cases at T1 showed a significant decrease in BMI compared to T0. The Δ T1 − T0 confirmed that these differences were significant in cases (Table 3).

Total, HDL, LDL cholesterol and triglycerides did not show any statistical difference between cases and controls at T0 and T1. The Δ T1 − T0 of fasting glucose was significant in cases. Insulin and C-peptide were higher in cases both at T0 and T1 in comparison to controls (Table 3).

Myeloperoxidase levels were significantly lower at time T1 than at time T0 in both controls and cases. The Δ T1 − T0 indicated that this difference was significant in cases (Figure 1). The amount of MDA, a marker of lipid peroxidation, was significantly lower at T1 both in controls and cases (Figure 2A). The production of 8-OHdG, a reliable marker of DNA oxidative stress, measured at T1, both in controls and in cases, was significantly decreased in comparison to the values at T0, mostly in cases (Figure 2B).

The levels of TNF-α and IL-6 showed a decreasing trend from T0 to T1, both in controls and cases, although these differences were not significant (Figures 3A, B). Interestingly, in cases, the level of IL-10, an anti-inflammatory cytokine, was significantly higher at T1 than at T0; no statistical significant differences were found in controls (Figure 3C).

At T0, the control group showed higher mean plasma adiponectin levels compared to those of cases. For both groups there was a statistically significant increase in adiponectin levels after MD rich in HQ-EVOO. T1 − T0 Δ resulted to be 0.6 ± 0.26 in controls while 1.6 ± 0.2 in cases (Figure 4).

In our study, the yield of dsDNA per gram of feces, extracted from all the samples, grouped as controls and cases at times T0 and T1, was not significantly different, neither between cases nor between times (controls T0/T1, cases T0/T1, controls T0/cases T0, controls T1/cases T1, data not shown). We can be assumed that there was no modification measurable by Qubit.

Regarding the specific LAB’s rpoB sequences quantified by qPCR, significant differences were assessed between controls and cases as a function of time. The rpoB copy number per ng of template DNA was significantly increased in cases and controls at T1; while, at T0 the rpoB copies between controls and cases were not significantly different (Figure 5). In particular, the amount of rpoB copies in controls at T1 increased 11.4 times than those in the same samples at T0, while in cases at T1 rpoB amount increased 55.6 times than those at T0. Moreover these results showed that the rpoB amount in cases at T1 were significantly higher than the ones in controls.