Gut bacterial composition
Microbiological analyses were performed on fecal samples to assess the growth of the HBF in germ-free mice and to ascertain the effects of probiotics on the development of gut bacteria. The measured terminal composition of the fecal microbiota is detailed in , where the statistically significant differences between the various groups were calculated using a two-tailed Mann–Whitney test. The bacterial populations of Bifidobacteria longum and Staphylococcus aureus were reduced after introduction of both probiotics. Additionally, unique effects of L. rhamnosus supplementation caused decreased populations of Bifidobacterium breve, Staphylococcus epidermidis and Clostridium perfringens but an increase of E. coli.
Microbial species counts in mouse feces at the end of the experiment
Gut levels of short-chain fatty acids
Short-chain fatty acids (SCFAs), namely acetate, propionate, isobutyrate, n-butyrate and isovalerate, were identified and quantified from the cecal content using GC-FID. The results, presented in , are given in μmol per gram of dry fecal material and as mean±s.d. for each group of mice. The production of some of the SCFAs, that is, acetate and butyrate, by the HBF mice supplemented with both of the probiotics was reduced. In addition, increases of the concentrations in isobutyrate and isovalerate were observed in the mice fed with L. paracasei.
Short-chain fatty acid content in the cecum from the different groups
Analysis of1H NMR spectroscopic data on plasma, urine, liver and fecal extracts
A series of pairwise O-PLS-DA models of 1
H NMR spectra were performed to extract information on the metabolic effects of probiotic modulation. A statistically significant metabolic phenotype separation between untreated mice and probiotic supplemented animals was observed as reflected by the high value of QY2
for each model (Cloarec et al, 2005b
; ). The corresponding coefficients describing the most important metabolites in plasma, liver, urine and fecal extracts that contributed to group separation are also listed in Supplementary Table 1
. The area normalized intensities (101
a.u.) of representative metabolite signals are given as means±s.d. in . The O-PLS-DA coefficients plots are presented in using a back-scaling transformation and projection to aid biomarker visualization (Cloarec et al, 2005b
). The direction of the signals in the plots relative to zero indicates positive or negative covariance with the probiotic-treated class. Each variable is plotted with a color code that indicates its discriminating power as calculated from the correlation matrix thus highlighting biomarker-rich spectral regions.
Summary of influential metabolites for discriminating NMR spectra of liver, plasma, fecal extracts and urine
Figure 1 O-PLS-DA coefficient plots derived from 1H MAS NMR CPMG spectra of liver (A, D), 1H NMR CPMG spectra of plasma (B, E), 1H NMR standard spectra of fecal extracts (C, F) and urine (G, H), indicating discrimination between HBF mice fed with probiotics (positive) (more ...) Liver metabolic profiles
Livers of mice fed with L. paracasei showed relative decreases in dimethylamine (DMA), trimethylamine (TMA), leucine, isoleucine, glutamine, and glycogen and increased levels of succinate and lactate (). Mice supplemented with L. rhamnosus showed relative decreases in leucine and isoleucine and relative increases in succinate, TMA and trimethylamine-N-oxide (TMAO) in the liver compared to controls ().
Plasma metabolic profiles
Plasma samples showed relative decreases in the levels of lipoproteins and increases in the concentrations of glycerophosphorylcholine (GPC) and triglycerides in mice fed with both probiotics compared to controls (). Elevated choline levels were observed in plasma of mice fed with L. rhamnosus and reduced plasma citrate levels were observed in mice fed with L. paracasei compared to controls.
Fecal extract metabolic profiles
Marked changes were observed in the metabolic profiles of fecal extracts from all supplemented mice, for example relative decreased concentrations of choline, acetate, ethanol, a range of putative N-acetylated metabolites (NAMs), unconjugated bile acids (BAs) and tauro-conjugated bile acids (). Furthermore, relative higher levels of glucose, lysine and polysaccharides were detected in the feces from mice fed with probiotics. A relative increased level of n-caproate (chemical shifts δ at 0.89(t), 1.27(m), 1.63(q), 2.34(t)) appeared to be associated with mice supplemented with L. paracasei.
Urine metabolic profiles
Urine samples of mice supplemented with both probiotics showed relative increased concentrations of indoleacetylglycine (IAG), phenylacetylglycine (PAG), tryptamine and a relative decrease in the levels of α-keto-isocaproate and citrate (). Relative increased concentrations of a mixture of putative glycolipids (UGLp, chemical shifts of multiplets at δ 0.89, 1.27, 1.56, 1.68, 2.15, 2.25, 3.10, 3.55, 3.60), N-acetyl-glycoproteins (NAGs) and a reduction in 3-hydroxy-isovalerate were also observed in mice supplemented with L. paracasei compared to controls. Urine of mice fed with L. rhamnosus showed a reduction in levels of creatine and citrulline.
UPLC-MS analysis of bile acids in ileal flushes
The proportion of bile acids in ileal flushes from the different groups are given in and are shown as mean±s.d. of the percentage of the total bile acid content. O-PLS-DA of the data set revealed that the relative concentrations of bile acids obtained from unsupplemented HBF mice are separated from those treated with probiotics, the correlation observed with L. paracasei being more significant than with L. rhamnosus as noted by the values of the cross-validated model parameter QY2 (). For example, HBF mice supplemented with L. paracasei showed strong correlations with higher amounts of GCA, CDCA and UDCA and lower levels of α-MCA in the ileal flushes when compared to controls. HBF mice fed with L. rhamnosus also showed higher levels of GCA associated with lower levels of TUDCA and TCDCA in the ileal flushes when compared to untreated HBF mice.
Bile acids composition in gut flushes for the different microbiota
Figure 2 O-PLS-DA coefficient plots derived from the bile acid composition obtained by UPLC-MS analysis of ileal flushes, which indicate discrimination between HBF control mice (negative) and HBF mice treated with probiotics (positive), (A) L. paracasei and ( (more ...)
Integration of multicompartment metabolic data using hierarchical-principal component analysis
A principal component analysis (PCA) model was initially constructed separately for the metabolic data from each individual biological matrix (plasma, urine, liver, fecal extracts and bile acid composition in ileal flush; and ). Three principal components were calculated for each cross-validated PCA model, except for the plasma where four principal components were retained to maximize the explained variance R2X
and the cross-validation parameter Q2
following the standard sevenfold cross-validation method (Cloarec et al, 2005b
). These PCA models descriptors (R2X
) were 0.70/0.36 (plasma), 0.41/0.11 (urine), 0.80/0.71 (liver), 0.94/0.67 (bile acid) and 0.61/0.42 (feces). The score vectors tb
from each model were then assigned as new X-variables (). Thus, the top level X-matrix contained 16 descriptors, denoted Pi
(plasma PCs 1–4), Li
(liver PCs 1–3), Ui
(urine PCs 1–3), Bi
(bile acid PCs 1–3) and Fi
(fecal PCs 1–3), which comprise only the systematic variation from each of the blocks/compartments. The two first principal components (p1 and p2) calculated for the hierarchical-principal component analysis (H-PCA; Westerhuis et al, 1998
) model (R2X
=0.60) accounted for 37 and 23% of the total variance in the combined multi compartment data respectively. The cross validation for the H-PCA model failed due to the high degree of orthogonality within the X-matrix, that is, within each of the blocks all variables are orthogonal to each other, while each of the biological matrices could be cross-validated at the individual level.
Figure 3 Schematic overview of H-PCA modelling: (Gunnarsson et al, 2003). In the sublevel, each block of data XB is modelled locally by a PCA model. Each block is summarized by one or more loading vectors pb and score vectors tp (‘super variables'), which (more ...)
Figure 4 H-PCA scores (A) and loadings (B) plots for the two first components derived from scores of separate PCA constructed separately for the metabolic data from each individual biological matrix from HBF mice (), HBF-L. paracasei mice () (more ...)
The H-PCA scores plot illustrated a degree of clustering with respect to the groups of HBF mice (). The corresponding H-PCA loadings plot indicated the contribution of the 16 descriptors to the differences observed between the samples in the H-PCA scores plot. Probiotic-supplemented HBF animals are separated from the controls along the first principal component, and this arises from the main variations modelled at the base level PCA of the individual plasma (P1, P2), liver (L2, L3), ileal flush (B2) and urine (U1, U2) data sets (). HBF mice fed with L. paracasei were separated from those fed with L. rhamnosus along the second principal component, which was mainly due to the variations modelled at the base level PCA of plasma (P2, P3), liver (L1) and urine (U1, U3). Interestingly, the metabolic variations in the fecal samples have no weight in discriminating the bacterial supplementation from the corresponding controls in the global model.
To uncover variables contributing to the H-PCA super scores, the loadings at the base level PCA model were interrogated (). HBF mice supplemented with probiotics showed higher concentrations of glucose, choline, GPC, glutamine, glutamate and lysine in the plasma profiles associated with elevated concentrations of glucose in the liver and higher levels of TUDCA and TCDCA in the ileal flushes. Controls showed higher levels of lipoproteins in the plasma, elevated concentrations of lipids, glycogen, glutamine, glutamate, alanine, TMAO and lactate in the liver, associated with higher levels of TCA and TβMCA in the ileal content. Unsupplemented HBF mice also showed elevated urinary excretions of creatine, citrate, citrulline, lysine, UGLp, NAG and α-keto-isocaproate compared to animals fed with probiotics. Moreover, H-PCA also revealed that HBF mice treated with L. rhamnosus had higher levels of hepatic lipids and plasma lipoproteins but lower concentrations of lactate and amino acids in plasma and lower urinary excretion of PAG, IAG, tryptamine and taurine than HBF mice treated with L. paracasei.
Integration of correlations between bile acids and fecal flora
We further investigated the connections between fecal flora and intestinal bile acids using a correlation analysis based bipartite graphical modelling approach (see Materials and methods; ) used previously to investigate the effects of gut microbiome humanization in germfree mice (Martin et al, 2007a
). In the current study, we are working with a superior model where all the major bacteria strains are identified, which was not possible when considering conventional microflora. Positive and negative correlations show the multicolinearity between bile acids and gut bacteria, whose concentrations are interdependent such as in the case of substrate–product biochemical reactions. Additional pixel maps of the correlation matrices are given to help interpretation in Supplementary Figure 1
Figure 5 Integration of bile acid and fecal flora correlations. The bipartite graphs were derived from correlations between fecal flora and bile acids in each group: HBF mice (A), HBF mice supplemented with L. paracasei (B) or L. rhamnosus (C). The cut-off value (more ...)
Control HBF mice and HBF mice supplemented with probiotics show remarkably different bile acid/fecal flora correlation networks (), indicating that small modulations in the species composition of the microbiome can result in major functional ecological consequences. Network statistics reveal that microbiome/metabolome bipartite graphs from HBF mice supplemented with Lactobacilli show a totally different nodal structure (given for the cut-off value of 0.5). In particular, we observed that in the network obtained from control HBF mice, the most connected bacteria are Bifidobacteria, Staphylococci, and Clostridia, for which the variations are intrinsically correlated with the balance in tauro-conjugated bile acids (TCDCA, TβMCA, TCA, TUDCA) and unconjugated bile acids (CA, αMCA, βMCA). In particular, the potentially harmful opportunist C. perfringens shows functional correlation of opposite sign for TβMCA, TCDCA, αMCA and βMCA when compared to Bifidobacteria and S. aureus.
Network analysis for HBF mice supplemented with L. paracasei reveals that Lactobacilli supplementation resulted in decreasing the functional links between Bifidobacteria and bile acids, while new significant correlations were observed between bile acids, and Bacteroides, S. aureus, S. epidermidis and L. paracasei. Moreover, E. coli has several connections with UDCA, αMCA and TCDCA. Interestingly, Bacteroides shows functional correlations of opposite signs for TβMCA, TUDCA and TCA when compared to Lactobacilli and Staphylococci.
When HBF mice received L. rhamnosus probiotic, the microbiome/metabolome network shows a significant lower level of complexity (given for the cut-off value of 0.5). The balance within B. breve, S. aureus and S. epidermidis appears highly correlated to the composition in tauro-conjugated bile acids (TβMCA, TCA) and unconjugated bile acids (βMCA, CA).