Animals
The animals for this experiment were pigs of similar genotype of Danish Landrace and Yorkshire. Six female siblings from a normal litter (the control group) (75% Landrace x 25% Yorkshire) were obtained after standard artificial insemination followed by caesarian section. The cloning experiments were performed using donor cells obtained from a 65% Landrace x 35% Yorkshire sow as described previously [9]. The cloned embryos were then transferred surgically to surrogate sows (recipients) five to six days after cloning [9]. Two surrogate sows gave birth to five live female clones by caesarean section. Pigs were reared in the experimental stables at University of Aarhus (Tjele, Denmark). All the experimental animal studies were approved by the Danish Animal Experimental Committee.
Experimental set up and sample collection
The pigs in the experiment were weaned at 28 days of age and subsequently fed a standard pig-diet with an energy distribution of 18.5% protein, 7.9% fat, 72.4% carbohydrate and 1.2% fiber, for approximately 61 days. During this post weaning period animals from the same litter were housed together in the same stable. At 96 days (cloned pigs) and 89 days (non-cloned controls) of age (baseline), the pigs were transferred to facilities for individual housing and fed a wheat-based HF/high-caloric diet consisting of 19.5% protein, 27% fat, 53% carbohydrates and 0.5% fiber [22] with ad libitum access to the feed in order to induce obesity. The feed was weighed before and after feeding and the pigs were maintained on this diet for a period of 136 days until they were euthanized. The cloned and non-cloned control pigs were weighed biweekly starting a day prior to switch to HF/high-caloric feed and the body-fat composition of the animals was measured by computed tomography (CT) scan at the end of the experiment. During this period, fresh feces collected biweekly were snap-frozen in liquid nitrogen and stored at −20°C until later analyses.
Terminal restriction fragment length polymorphism (T-RFLP)
The fecal microbiota from all the pigs were analyzed by terminal restriction fragment length polymorphism (T-RFLP) fingerprint profiles as described previously [23]. In brief, DNA was extracted from 200 mg feces by using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions, with an additional step of bead beating in order to disrupt the cell wall of Gram-positive bacteria. The concentrations of DNA were measured in each sample by a spectrophotometer and adjusted to 5 ng μl-1 (NanoDrop Technologies,Wilmington, DE, USA). Amplification of 16S rRNA gene DNA were performed in duplicates by using 16S rRNA gene DNA bacterial specific primers, Eub-8fm (5’- AGAGTTTGATCMTGGCTCAG- 3’) labeled with 5´ FAM and Eub-926r (5-’CCGTCAATTCCTTTRAGTTT- 3’) (DNA Technology, Aarhus, Denmark) [23]. Each PCR mix contained 5 μl of 10x Fermentas Taq-buffer, 4 μl MgCl2, 2.0 μl deoxyribonucleotide triphosphate (dNTP), 0.5 μl Fermentas Taq-polymerase, 0.5 μl of each primer and 35.5 μl nuclease-free water and 5 ng μl-1 DNA (final concentration of 0.2 ng). The cycling conditions were: initial denaturation at 94°C for 6 minutes (min) followed by 32 cycles of denaturing at 94°C for 45 seconds (s), annealing at 56°C for 45 s, an extension step at 72°C for 2 min, and a final extension at 72°C for 10 min. The PCR products were subsequently verified by gel electrophoresis and purified by High Pure PCR Purification Kit (Roche Applied Sciences, Mannheim, Germany). The purified PCR product (200 ng) was digested with 2.0 μl of the restriction enzyme HhaI (Promega Corporation, Madison, USA) at 37°C for 3 h. Two μl of the digested PCR products, 10 μl formamide and 0.50 μl Megabase ET900-R Size Standard (GE Health Care, Buckinghamshire, UK) were mixed and run in duplicates on a capillary electrophoresis genetic analyzer (Genetic Analyzer 3130/3130xl, Applied Biosystems, Carlsberg, CA). The terminal restriction fragments (T-RFs), representing bacterial fragments in base pair (bp), were obtained and the analysis of T-RF profiles and alignment of T-RFs against an internal standard was performed using the BioNumerics software version 4.5 (Applied Maths, Kortrijk, Belgium).
T-RF fragments (range of 60–800 bp) with a difference less than two base pairs were considered identical. Only bands present in both duplicates were accepted as bacterial fragments from which the duplicate with the best intensity was chosen for microbial profiling. The obtained intensities of all T-RFs were imported into Microsoft Excel, and all intensities below 50 were removed. In each sample, the relative intensity of any given T-RF was calculated by dividing the intensity of the T-RF with the total intensity of all T-RFs in the sample. The most predominant T-RFs with a mean relative intensity above one percent were selected for all further analyses and procedures (except calculation of the diversity and similarity) and their identity was predicted in silico, performed in the MiCA on-line software [24] and Ribosomal Database Project Classifier (322.864 Good Quality, >1200) [25].
T-RFLP statistical analysis
All T-RFs between 60 and 800 bp were imported into the statistical software programs Stata 11.0 (StataCorp, College Station, TX), Unscrambler version 9.8 (CAMO, Oslo, Norway) and Microsoft Excel sheets were used for further analyses. Principal component analysis (PCA) was used to explore group differences in the overall microbial communities both for comparisons between cloned pigs and non-cloned controls at the different sampling points and to investigate if samples from pigs with the largest weight-gain during the study period clustered together, irrespective of their genetic background. The latter was also investigated by relating the whole microbial community to the weight-gain at the different sampling points, involving all predominant T-RFs simultaneously in the models. For this purpose partial least square regression (PLS-R) was used, which is a supervised model, meaning in this case that the variation in the weight (gain) data is used to actively decompose the variation in the bacterial data. In both analyses, the T-RFs were standardized (centered and 1/SD) prior to the modeling phase to ensure that all of them would equally influence the models, and possible outliers were inspected visually and with Hotelling T2.
The diversity index was calculated as described previously [26]. In brief, the Shannon-Weaver index of diversity (H’) based on all of the initial T-RFs was used to determine the diversity of the bacterial fragments. Group comparisons of the diversity index in cloned versus non-cloned controls were calculated at each of the sampling points. As the Shannon-Weaver index was not normally distributed, Mann Whitney U test and Spearman correlation were applied. The H’ values are represented in figures as mean and error bars representing standard deviations (SD). Dice similarity between groups based on all the T-RFs were calculated in BioNumerics (Applied Maths, Kortrijk, Belgium) and the results are presented as mean values. T-RFs in the figures are presented as mean and standard error of the mean (SEM). A significant difference was considered when P-value was less than 0.05 (P<0.05).
Fecal samples and bacterial strains for qPCR
The extracted DNA from the fecal samples used for the T-RFLP analyses were also analyzed by qPCR, but only samples taken monthly were chosen for qPCR analysis. However additional sampling points two weeks before the endpoint samples were also analyzed by qPCR. Three bacterial strains (Clostridium perfringens (NCTC 8449), Odoribacter splanchnicus (isolate DJF_B089) and Escherichia coli (ATCC 25922), representing the Firmicutes and Bacteroidetes phyla and general bacteria, respectively, and six randomly chosen extracted DNA samples (divided equally into clones and controls) were used to optimize the PCR conditions.
qPCR primers and conditions
The 16S rRNA gene DNA primers for Bacteroidetes and Firmicutes used in this study were designed by Baccetti De Gregoris et al.[27] and conditions were optimized for the thermocycler used (Rotor-Gene Q Real Time PCR cycler (Qiagene)). The universal primer used in this study had an amplicon length of 147 bp (S-D-Bact-0907-a-S-20 5’-AAACTCAAAGGAATTGACGG-3’; S-D-Bact-1054-a-A-20 5-’ ACGAGCTGACGACAGCCATG-3’) [12]. The specific primer sets for Bacteroidetes (798cfbF 5’ CRAACAGGATTAGATACCCT’3 and cfb967R 5’ GGTAAGGTTCCTCGCGTAT ‘3) and Firmicutes (928F-Firm 5’ TGAAACTYAAAGGAATTGACG ‘3; 1040firmR, 5’ ACCATGCACCACCTGTC ‘3) had an amplicon length of 240 bp and 200 bp, respectively [27]. All qPCR reactions contained 12.5 μl of SYBR® Green JumpStart™ Taq ReadyMix™ without MgCl2 (Sigma-Aldrich, Copenhagen, Denmark), 0.3 μmol l-1 of each primer and 5 μl of template DNA adjusted to 5 ng μl-1. MgCl2 optimization was performed and a final concentration of 2.5 mM MgCl2 was chosen. The annealing temperature was optimized by using 16S rRNA gene DNA extracted from fecal samples and DNA extracted from different bacteria. Subsequently, all the primers and other PCR conditions were verified by conventional PCR and gel electrophoresis. A non template control (NTC) was included in each run. qPCR was performed with an initial denaturing step of 10 min at 95°C, 95°C for 30 s, 35 cycles of 56°C for 20 s and an elongation step of 72°C for 20 s. A melting curve analysis was performed after each run to detect any primer-dimers in each sample. The threshold cycle (C
T
) and calculated concentrations (copies μl-1) were determined automatically by the Rotor Gene software (Rotor-Gene Q 2.0.2 (Qiagene)).
Analysis of data from qPCR
qPCR was performed to quantify relative abundance of the phyla Bacteroidetes and Firmicutes, respectively, present in each sample. The measured bacterial copy numbers of the 16S rRNA gene from bacteria belonging to the phylum Bacteroidetes and the phylum Firmicutes were calculated against 16S rRNA genes obtained from all bacteria and the relative abundance of the two phyla in each sample was subsequently calculated and statistically evaluated by Mann Whitney U test. Further correlation analyses were performed using Spearman correlation coefficient and P <0.05 was considered statistically significant. A standard curve was constructed for specific and universal primer sets and assays using tenfold serial dilutions of the extracted DNA from C. perfringens, O. splanchnicus and E. coli all DNA samples in the range 2.5 x102 ng μL-1 to 2.5x10-6 ng μL-1. Furthermore, serial dilutions corresponding to the previously described dilutions of genomic DNA from two random samples were used to construct standard curves to further verify if PCR inhibitors were present in extracted DNA from fecal samples.