Resveratrol, lunularin and dihydroresveratrol do not act as caloric restriction mimetics when administered intraperitoneally in mice

Resveratrol as well as caloric restriction were shown to extend lifespan in some model organisms and may possibly delay onset of ageing-related diseases in humans. Yet, resveratrol supplementation does not always extend lifespan of animal models or improve health status of humans. Because of interindividual differences in human microbiota, resveratrol metabolite production in the gut differs. While some individuals produce lunularin and dihydroresveratrol in their gut, others produce dihydroresveratrol only. Therefore, we addressed the question whether these metabolites differ in their biological impact on ageing and intraperitoneally injected 13-month-old C57BL/6JRj mice on an ad-libitum (AL) HFD with resveratrol, dihydroresveratrol or lunularin (24 mg/kg bodyweight; 3 times/week). Compared to mice injected with vehicle (AL-control), resveratrol and dihydroresveratrol did not change bodyweight and had no impact on insulin or glucose levels while lunularin slightly reduced feed intake and bodyweight gain. CR-mice showed lowered cholesterol, insulin and leptin levels, elevated adiponectin and phosphorylated AMPK levels in liver as well as increased transcription of Pck1 and Pgc1α when compared to the AL-control. In contrast, injections with the test substances did not change these parameters. We therefore conclude that in our model, resveratrol, lunularin and dihydroresveratrol did not act as CR mimetics.


Supplemental Data X-1
Mouse weight and average feed intake during the injections with resveratrol, dihydroresveratrol, lunularin or DMSO/saline. Apart from the CR mice, the animals had free access to feed. (500 pmol/g liver, n = 6; SD, standard deviation)

Linearity
A best fit line was obtained by linear regression using a weighting of 1/x². The correlation coefficient was used as an indicator for the quality of the calibration curves and was >0.9944.

Limit of quantitation (LOQ) and limit of detection (LOD)
LOQ and LOD were defined as the levels with a signal-to-noise (S/N) ratio of 10 and of 3, respectively.

Analysis of resveratrol and metabolites in mouse liver samples
The UHPLC-MS/MS analysis and the sample preparation for these analyses took place at the MRI, Karlsruhe, Germany. Resveratrol (>99%) for UHPLC-MS/MS analyses was purchased from Sigma Aldrich while dihydroresveratrol (>99.5%) was purchased from LKT Laboratories (St. Paul, Minnesota, USA).
All other chemicals and solvents used were of analytical grade. Deionized water was taken from an inhouse ultrapure water system (LaboStar; Siemens, Erlangen, Germany) with a conductivity of 0.055 μS/cm.
For sample preparation mouse livers were homogenized with a ball mill (MM400; Retsch, Haan, Germany). Before starting the homogenization, liver samples and the grinding beakers including grinding balls were pre-cooled with liquid nitrogen. Homogenized liver samples were portioned in approx. 100 mg aliquots and stored at -80°C until analysis. The sample preparation procedure was done on ice wherever possible. Quantification of the analytes was done by external calibration.
Therefore analyte-free blank mouse liver samples were spiked with 5 µL of internal standard (10 µM of 13 C6-resveratrol in DMSO) and 5 µL of analyte standard solutions in DMSO (resveratrol, dihydroresveratrol and lunularin each in an end level range between 4 and 2500 pmol/g liver).
Aliquots of mouse liver samples were spiked with 5 µL of internal standard ( Germany). To obtain the calibration curves, the ratios of the analyte peak areas to the internal standard peak area were calculated and plotted against the analyte level. A best fit line was obtained by linear regression using a weighting of 1/x 2 .
The method was validated based on FDA criteria. In detail, selectivity, accuracy, intra-day precision, recovery, linearity, limit of quantitation (LOQ) and limit of detection (LOD) were determined. The validation results are summarized in Supplemental Data X-3.

Statistics
The statistical software R 36 was used to evaluate the data. Appropiate models were defined 37,38 . This was a linear model for the blood parameters, feed intake, blood parameters and Ussing chamber X6 measurements and a mixed model for the PCRs. Here, the treatment group was regarded as a fixed factor and the PCR run was regarded as a random factor. For the weekly weight development, the statistical model included the treatment group (CON, RSV, DHR, LUN), the covariate 'week' as well as their interaction term as fixed factors. The mouse was regarded as random factor. We excluded the CR mice from the statistical analysis of feed uptake and bodyweight.
The data was assumed to be approximately normally distributed. For the weight development, the data was assumed to be heteroscedastic due to the week and treatment group. These assumptions are based on a graphical residual analysis. Based on this model, a pseudo R 2 was calculated 62 . While for the blood parameters, feed intake and normalized mRNA levels an analysis of variances (ANOVA)