Growth of Bifidobacterium species is inhibited by free fatty acids and bile salts but not by glycerides.

High-fat diets have been associated with lower gut and fecal abundances of genus Bifidobacterium. Here, we investigated whether commonly consumed dietary free fatty acids have any detrimental effect on the growth of B. adolescentis, B. bifidum, and B. longum. We found that the presence of free fatty acids in the medium inhibits the growth of Bifidobacterium species to a varying degree, with capric (C10:0), oleic (C18:1), and linoleic (C18:2) acids displaying the largest effect. In comparison, free fatty acids did not affect the growth of Escherichia coli. When fats were added as a mixture of mono- and diacylglycerols, the inhibitory effect on Bifidobacterium growth was abolished.

Dietary fats comprise a major part of human diet. Nutrient recommendations call for fats accounting for 20–35% of the daily calories [1]. The actual consumption of fats in western countries is significantly higher, providing around 40% of caloric intake [2]. While dietary fats are efficiently absorbed in the small intestine [3], increased dietary intake of fats likely saturates its absorption capacity [4], and so substantial amounts of lipids reach the colon along with some bile salts. In the colon, these compounds interact with resident gut microbiota [5],[6]. Recently, using an in vitro Human Gut Simulator system, we showed that human gut microbiota can efficiently utilize dietary free fatty acids for growth, and we identified specific “lipidophilic” and “lipidophobic” members [7]. Surprisingly, Bifidobacterium members were found in unexpectedly low numbers in all samples, even in the balanced medium that contained abundant carbohydrate and protein nutrients in addition to free fatty acids. Abundance of this genus was also reported to be low in several studies of high-fat diets, both in humans [8],[9] and in animal models [10].

The goal of this study was to determine if fatty acids, commonly consumed as part of human diet, and/or bile salts could have an inhibitory effect on the growth of several Bifidobacterium species. The most abundant colonic species of free fatty acids [7] were selected for testing and included capric (C10:0), palmitic (C16:0), stearic (C18:0), oleic (C18:1), and linoleic (18:2) acids. Three Bifidobacterium species with well-established prevalence in human gut (Bifidobacterium longum subsp. longum DSM-20219, Bifidobacterium bifidum DSM-20456, Bifidobacterium adolescentis L2-32) were grown in either DSMZ Medium 58 (“Bifidobacterium medium”, BM) as recommended by the German Collection of Microorganisms and Cell Cultures, or in the balanced Western diet medium (WM), which we used previously to mimic food bolus contents reaching colon in subjects consuming a typical Western diet [7]. Escherichia coli K12 NCM3722 [11], which was previously found to tolerate well the presence of fats in the growth medium [12], was used as a control, and was cultured in the Western diet medium.

The experimental design was as follows: each Bifidobacterium strain was cultured in both BM and WM media with or without fatty acids and/or bile salts. Control medium contained no fatty acids or bile salts. Fatty acids were tested individually as well as in mixture in concentrations matching the contents of the Western diet medium, which in turn was developed to match bolus contents reaching the colon [7]. Medium composition is detailed in Table 1. A commercially available mixture of mono- and diacylglycerols (Modernist Pantry, LLC) was also used. All cultures were grown in Hungate tubes at 37 °C under fully anaerobic conditions. Atmosphere was 85% of N2, 10% of CO2, and 5% of H2. All Hungate tubes were inoculated to the same starting density from a single pre-culture tube. All experiments were performed in triplicates. After 24 hours of growth, cell densities of each culture were obtained via phase contrast microscopy with a Spencer hemocytometer as we did previously [13]. Statistical significance of culture density differences between samples was assessed with Student's t-test at a confidence level of 95%. Raw p-values were adjusted for multiple hypothesis testing according to Benjamini and Hochberg method [14].

Table 1.

Medium composition, g l−1.

Medium component	BM	BM with added fats	WM control	WM
Carbohydrates
Arabinogalactan	-	-	1.8	1.8
Guar gum	-	-	0.9	0.9
Inulin	-	-	0.9	0.9
Pectin	-	-	1.8	1.8
Starch	-	-	4.4	4.4
Xylan	-	-	0.9	0.9
Cellobiose	-	-	0.9	0.9
Glucose	10	10	0.5	0.5
Fructose	-	-	0.5	0.5
Proteins
Peptone	10	10	3.3	3.3
Casein	-	-	2.0	2.0
Lipids
Capric acid (C10:0)	-	0.3	-	0.3
Palmitic acid (C16:0)	-	1.5	-	1.5
Stearic acid (C18:0)	-	0.7	-	0.7
Oleic acid (C18:1)	-	1.8	-	1.8
Linoleic acid (C18:2)	-	1.2	-	1.2
Mucin	-	-	4.0	4.0
Yeast extract	5.0	5.0	3.0	3.0
Meat extract	5.0	5.0	-	-
Soy extract	5.0	5.0	-	-
Salts, other components	22.5	22.5	14.9	14.9
Bile salts	-	1.0	-	1.0

BM–Bifidobacterium medium; standard medium contained no fatty acids or bile salts.

WM–Western diet medium; control medium contained no fatty acids or bile salts.

All three tested Bifidobacterium species reached stationary phase after 24 hours of incubation. While there was some variability in the final cell densities reached, the values were consistent for each species across replicate cultures. All three species experienced inhibition by fatty acid pool (containing five different free fatty acids), and B. bifidum and B. longum were also significantly inhibited by bile salts (Figure 1). Overall, B. longum and B. bifidum were similarly affected across tested conditions. The presence of capric acid (C10:0), oleic acid (C18:1), and linoleic acid (18:2) had statistically significant negative effect on culture growth, whereas palmitic (C16:0) and stearic (C18:0) acids had no effect. On the other hand, B. adolescentis was not inhibited by bile salts but was significantly affected by each of the five tested free fatty acids. Oleic and linoleic acids have been previously observed to be inhibitory towards Gram-positive but not Gram-negative bacteria [12] including inhibition of B. breve [15]. Exposure to linoleic acid was found to alter B. breve metabolism triggering an increased oxidative stress that resulted in the lower growth rate [15]. Generally, free fatty acids possess membrane-destabilizing activity due to their amphipathic structure [16], and unsaturated fatty acids can also elicit bacteriostatic properties through the inhibition of fatty acid synthesis [12]. However, no inhibitory effect on Bifidobacterium was found in another investigation that tested the antimicrobial effect of different oils, all containing oleic and linoleic acids [17].

Figure 1.

Cell densities of Bifidobacterium stationary phase cultures grown in Bifidobacterium medium. “Fats” denote a mixture of five tested fatty acids. Statistical significance was obtained via Student's one-tail t-test with α = 0.05 and multiple hypothesis testing adjustment. Statistical comparisons were made using the control as reference group. Significance labels: *: q < 0.05; **: q < 0.01; ***: q < 0.001; ns: not significant.

In our experiments, it was capric acid that showed the largest inhibitory effect across all three Bifidobacterium strains (Figure 1). To the best of our knowledge, capric acid's inhibitory effects on Bifidobacterium have not been tested previously, though it was shown to be inhibitory to other bacteria, possibly due to the disruption of ion transport through the membrane [18]. Interestingly, the concentration of capric acid in breast milk and breastfeeding formula was negatively correlated with the total relative abundance of Bifidobacterium in infants' fecal microbiota [19]. On the other hand, palmitic and stearic acid did not affect the growth of either B. longum or B. bifidum, which is consistent with other studies that found no harmful effects of saturated long-chain fatty acids on certain Gram-positive bacteria [12]. Surprisingly, our observations show that both of these fatty acids are able to hinder the growth of B. adolescentis, which had not been described previously.

Bile salts are also known to possess antimicrobial properties due to their ability to alter cell membrane and to cause DNA damage [20], and the amount of bile salts released into the intestinal lumen correlates with the total fat content in consumed foods [21]. Thus, we also tested whether bile salts (a mixture of sodium glycocholate and sodium taurocholate) can have similar inhibitory effect on the Bifidobacterium growth. Tested Bifidobacterium species differed in their response to bile salts presence in the BM: bile had the largest inhibitory effect on the growth of B. longum, statistically significant inhibition of B. bifidum, and no statistically significant effect on the growth of B. adolescentis. Bifidobacterium strains that inhabit the gut may have developed various mechanisms to overcome the inhibition by bile acids. These adaptations include efflux of bile salts from cells, bile salt hydrolysis, and structural and compositional changes of bacterial cell membranes [20],[22]. All three tested Bifidobacterium species indeed encode bile salt hydrolase (EC:3.5.1.24, KEGG orthology: K01442) in their genomes.

In human gastrointestinal tract, majority of fatty acids reach the colon as partially hydrolyzed mono- or diacylglycerols. Therefore, we also tested an ability of a commercial mixture of mono and diacylglycerols (“glycerides”) to promote or inhibit Bifidobacterium growth. According to the manufacturer, this mixture can contain esterified caprylic (C8:0), capric (C10:0), lauric (C12:0), myristic (C14:0), palmitic (C16:0), stearic (C18:0), and oleic (C18:1) acids, with linoleic (C18:2) and linolenic (C18:3) acids being present in lesser amounts. This glyceride mixture did not exert any significant effect on any of the Bifidobacterium strains when they were cultured in BM. Because free hydroxyl groups of glycerol lower overall hydrophobicity of partially esterified acylglycerols and increase their polarity, this might lower their inhibitory effect on Bifidobacterium growth via decreased interactions with bacterial cell membranes.

Next, we carried out the same set of experiments using Western diet medium that has been developed to maintain complex human gut microbiota in the in vitro Human Gut Simulator system [7]. The results of this set of experiments generally matched those observed for the BM (Supplementary Figure S1). Specifically, all three Bifidobacterium species were significantly inhibited by capric, oleic, and linoleic acids. Bile salts were only inhibitory to B. longum. The mono- and diacylglycerol mixture had no detrimental effect. Comparing the three Bifidobacterium species, B. bifidum was the most resistant to growth inhibition across different testing conditions.

Because Escherichia and total Gammaproteobacteria were found to increase on fats-only medium in our recent study [7], we used this genus of Gram-negative bacteria as a resistant control. E. coli was cultured in WM under the same experimental setup. Free fatty acids and their mixtures had no significant inhibitory effect on E. coli growth (Table 2), which is congruent with our previous observations [7] and those of other studies [12]. The observed resistance can likely be attributed to the outer membrane of Gram-negative bacteria being a barrier to hydrophobic compounds [23].

Table 2.

Cell densities of E. coli stationary phase cultures.

Medium composition	Cells ml−1
WM control	1.10 x 109 ± 1.83 x 108
WM + fatty acids and bile salts	1.07 x 109 ± 7.64 x 107
WM + fatty acids	0.97 x 109 ± 5.91 x 107
WM + bile salts	1.12 x 109 ± 8.13 x 107

WM-Western diet medium

Data are shown as arithmetic mean (n = 3) ± standard deviation

In conclusion, our results indicate that the presence of some free fatty acids in the environment, specifically capric, oleic, and linoleic acids, can hinder Bifidobacterium growth. Bile salts also inhibited Bifidobacterium growth but to a lesser degree. The mixture of mono- and diacylglycerols had no detrimental growth effect. We propose that when dietary fats are provided to gut microbiota community as part of growth medium, the free fatty acids are substituted with acylglycerols to mitigate any inhibitory effects on specific microbiota members.

Click here for additional data file.