Exploring diet effects on the profile, excretion and enterohepatic circulation of bile acids in the chicken

Abstract

Beyond their role in lipid digestion, the physiological function, distribution and composition of bile acids (BAs) in poultry remain insufficiently understood. We investigated bile acid metabolism in broiler chickens as well as laying hens in response to different diets. Two experiments were conducted: the first was a digestibility study of both genotypes fed either a highly digestible or challenge diet to assess the effect on fat digestibility and fecal bile acid excretion. In broilers, the highly digestible corn-soy diet improved fat digestibility and increased plasma triglycerides and non-esterified fatty acids (NEFA), whereas laying hens showed limited dietary effects. The bile acid excretion correlated only weakly with fat digestibility in both species. In the second experiment with broilers, BAs were measured in different matrices: gall bladder, intestinal content and mesenteric and peripheral blood pools. All pools were dominated by (tauro)chenodeoxycholic and taurocholic acid. Intestinal reabsorption of the conjugated BAs was far more efficient than the deconjugated BAs. Diet did not affect intestinal or circulating bile acid concentrations. Bile acid concentrations in peripheral circulation resembled the concentrations in mesenteric veins. In birds, fat digestion is less likely driven by the enterohepatic cycle. Instead, BAs appear to circulate more openly in peripheral circulation indicating that avian bile acid metabolism could be more complex than in mammals and cannot alone explain diet-induced differences in fat digestibility.

Introduction

Among avian species, chickens are the most studied (Flores-Santin and Burggren, 2021). Their digestive physiology has been thoroughly characterized to optimize nutrient digestibility and intestinal health, making the chicken a valuable model for investigating avian digestive function. Enzymes such as proteases, peptidases, carbohydrases, and pH have been characterized as driving factors to break down proteins and carbohydrates in aqueous media. Lipases, needed to hydrolyze triglycerides, are facilitated by emulsifying agents such as bile acids (BAs). The facilitation of fat digestion by BAs and other emulsifiers makes them potential feed additives in poultry feed (Ravindran et al., 2016; Zaefarian et al., 2019). Because of convergent evolution between birds and mammals it has been assumed that bile salt circulation and functionality is similar to mammalian monogastrics. Apart from fractional composition (Hagey et al., 2010; Tancharoenrat at al., 2022), recent studies on the metabolism of BAs in domesticated birds are scarce.

BAs are more functional when conjugated, in birds almost exclusively with taurine (Hagey et al., 2010). Next to this, once the BAs reach the intestinal tract, bacteria can deconjugate and transform BAs into secondary BAs, reducing their emulsifying action leading to altered performance (Foley et al., 2019). Exploring bile acid abundance and conjugation ratio in the intestine and body of chickens will yield further understanding of the avian metabolism of BA. Moreover, by including both growing broiler chickens, with a metabolism focused on protein accretion, and adult laying hens with high demanding fat metabolism due to yolk production, we get a holistic view on the different physiological states of chickens.

The first experiment aimed to exert changes in bile acid metabolism in broiler chickens as well as laying hens, by means of two diets with different carbohydrate, protein and fat source, leading to differing nutrient digestibility. A second experiment assessed bile salt concentrations in various matrices within the broiler chicken fed two types of diet as a mean to challenge metabolism to understand its effect on reabsorption and the enterohepatic cycle.

Materials and methods

All experimental procedures in this study were in compliance with the European guidelines for the care and use of animals in research (Directive 2010/63/EU) and were approved by the Ethical Committee of the Research Institute for Agriculture, Fisheries and Food (ILVO), Merelbeke, Belgium under authorization numbers 2018/321, 2018/350, 2022/406.

Animal trial 1

This trial consisted of two parts. In a first part of the trial, 108 male broilers (Ross 308) were purchased from a local hatchery (Belgabroed, Merksplas, Belgium). At arrival, chicks were weighed and divided in 36 pens. The chickens were reared on solid floor and wood shavings. A 18L6D lighting schedule was used and temperature of the stable was 32°C in the first week after which it declined gradually to 22°C in the fourth week. This temperature was maintained for the rest of the trial. Feed and water were offered ad libitum.

The birds were offered either of two basal diets (n = 18): a corn-soy based diet (CS) and a wheat-rapeseed-palm oil based diet (WRP) (Table 1). The rationale behind the diet challenge was to exert differences in fat digestion and metabolism. Diets were formulated to meet the birds’ requirements, and both diets were isocaloric and isonitrogenous. The 17-day-old birds were allocated to 36 digestibility units. Each digestibility unit housed three birds and was considered as one replicate.

Table 1.

Diet composition at the moment of sampling.

	Trial 1				Trial 2
	Broiler (Grower)		Layer		Broiler (Finisher)
Ingredient (%)	CS	WRP	CS	WRP	Control	Wheat
Corn	63.50	-	52.10	-	25.69	-
Wheat	-	64.60	5.0	57.30	43.09	69.54
Wheat bran	1.86	-	-	-	-	-
Soybean meal (48% CP)	28.10	14.00	19.10	4.50	16.90	8.78
Full fat soybeans	-	7.50	5.00	8.00	7.50	8.60
Rapeseed meal	-	6.50	-	5.00	-	6.00
Sunflower seed meal	-	-	-	10.00	-	-
Alfalfa	-	-	5.00	-	-	-
Lard	1.50	1.50	-	-	-	1.5
Soy oil	1.58	-	2.50	-	3.54	-
Palm oil	-	2.40	-	4.70	-	2.5
Vitamin & mineral premix1	1.00	1.00	1.00	1.00	1.00	1.00
CaCO3	0.620	0.560	8.900	8.600	0.500	0.315
Di-Ca-phosphate	0.720	0.650	0.683	0.154	0.615	0.490
NaCl	0.190	0.150	0.236	0.235	0.187	0.166
Na2CO3	0.220	0.330	0.220	0.105	0.188	0.202
L-Lysine HCl	0.261	0.310	0.001	0.151	0.282	0.383
DL-methionine	0.295	0.340	0.176	0.131	0.226	0.215
L-threonine	0.090	0.100	0.074	-	0.136	0.147
Calculated nutrient composition
Crude protein (g/kg)	195	194	166	166	185	185
Crude fat (g/kg)	70	70	63	75	73	74
Crude fibre (g/kg)	33	36	41	47	29	32
Metabolizable energy (kcal/kg)	2 844	2 820	2 844	2 844	2964	2904
Dig. Lysine (g/kg)	10.0	10.1	7.0	7.0	9.9	9.8
Dig. Methionine+Cysteine (g/kg)	7.8	8.9	5.8	5.9	7.4	7.4
Dig. Threonine (g/kg)	6.5	6.5	5.7	4.7	6.6	6.5
Dig. Valine (g/kg)	7.3	7.4	6.5	6.4	7.7	6.8
Ca (g/kg)	7.5	7.5	37.5	35	6.5	6.5
Available P (g/kg)	3.5	3.5	3.0	2.5	3.3	3.3
Na+Cl-K (meq/kg)	225	223	224	174	197	183
C18:2 (g/kg)	26.9	18.6	30.8	17.3	17.3	18.9
C18:1 (g/kg)	18.0	21.8	13.6	22.5	7.2	21.8
C16:0 (g/kg)	9.1	17.3	6.6	21.7	3.9	17.8

^1,layer Vitamin and mineral premix composed of vitamin A/retinyl acetate (10000 IU/kg); vitamin D3 (2999 IU/kg); vitamin E (all-rac-alpha-tocopheryl acetate) (50 IU/kg); vitamin K3 (2.5 mg/kg); vitamin B1/thiamine mononitrate (2 mg/kg); vitamin B2/riboflavin (5 mg/kg); calcium d-pantothenate (15 mg/kg); vitamin B6/pyridoxine hydrochloride (4 mg/kg); vitamin B12/cyanocobalamin (0.25 mg/kg); niacinamide (30 mg/kg); folic acid (1 mg/kg); biotin/D-(+)-biotin (0.15 mg/kg); choline chloride (690mg/kg); iron(II)sulphate (monohydrate) - iron (49 mg/kg); copper(II)sulphate (pentahydrate) - copper (20 mg/kg); zinc oxide (60 mg/kg); manganese(II)oxide - manganese (96 mg/kg); calcium iodate (anhydrous) - iodine (1.20 mg/kg); sodium selenite - selenium (0.36 mg/kg); sepiolite (7 mg/kg); propyl gallate (2 mg/kg); BHT (3 mg/kg).

^1,broilerVitamin and mineral premix composed of vitamin A/retinyl acetate 3a672a (1000000 IU/kg); vitamin D3 E671 (299999.4 IU/kg); vitamin E 3a700 (all-rac-alpha-tocopheryl acetate) (5000 IU/kg); vitamin K3 3a710 (250 mg/kg); vitamin B1/thiamine mononitrate 3a821 (200 mg/kg); vitamin B2/riboflavin (500 mg/kg); calcium d-pantothenate 3a841 (1500 mg/kg); vitamin B6/pyridoxine hydrochloride 3a831 (400 mg/kg); vitamin B12/cyanocobalamin (2.5 mg/kg); niacinamide 3a315 (3000 mg/kg); folic acid 3a316 (100 mg/kg); biotin/D-(+)-biotin 3a880 (15 mg/kg); choline chloride 3a890 (68965.5 mg/kg); iron(II)sulphate (monohydrate) - iron E1 (4920 mg/kg); copper(II)sulphate (pentahydrate) - copper E4 (2000 mg/kg); zinc oxide 3b603 (6000 mg/kg); manganese(II)oxide - manganese E5 (9590.2 mg/kg); calcium iodate (anhydrous) - iodine 3b202 (120 mg/kg); sodium selenite - selenium E8 (36 mg/kg); sepiolite E562 (700 mg/kg); propyl gallate E310 (200 mg/kg); BHT E321(300 mg/kg); citric acid.

Similar to the broiler part, twenty-four 32-week-old brown laying hens (ISA Brown) were purchased from a commercial farm and given the experimental feed for two weeks. A similar rationale for the two experimental diets (n = 12) was maintained for laying hens and formulated according to their requirements (Table 1). Subsequently, the birds were individually allocated in similar units as the broilers following the same protocol.

Digestibility of fat was measured according to the European reference method and started after an adaptation period of 4 days (Bourdillon et al., 1990). Total feed consumed and total amount of produced excreta were weighed to calculate feces to feed ratio. The excreta were homogenized and stored at −18°C in air-tight containers. Subsequently, samples were freeze-dried, ground to 1 mm and stored at ambient temperature for further analysis. Dry matter (152/2009/EG), crude protein (ISO 5983-2) and crude fat (ISO 6492) were analyzed using ISO standards from which apparent digestibility coefficients (digestibility coefficient crude protein (DCCP); digestibility coefficient crude fat (DCCF)) were calculated using the feces to feed ratio. Crude protein digestibility was corrected for fecal uric acid (Marquardt, 1983).

Bile salt concentration in freeze-dried finely ground excreta (<1 mm) was analyzed using the method described below. The concentration of BAs in the freeze-dried excreta was corrected for dry matter and multiplied by the total daily produced excreta during the digestibility trial.

Animal trial 2

24 male broilers (Ross 308) were purchased from a local hatchery (Belgabroed, Merksplas, Belgium). The animals were housed in a similar way as in trial 1 and divided in 4 pens. Again, two types of feed (Control and Wheat) were formulated according to the standards of trial 1 (Table 1). Feed contained 0.4% TiO₂ as inert marker to study within-intestine reabsorption ratio.

On day 39, six broilers per pen were sampled. Peripheral blood was sampled from the brachial vein to collect serum. Subsequently, animals were euthanized with pentobarbital IV (Euthanimal 200 mg/ml, Alfasan, Woerden, The Netherlands) via the metatarsal vein. Intestinal contents of the duodenum and colon were from the sampled non-fasted animals (Fig. 1). A needle was inserted in the gall bladder to aspire the bile without blood contamination. Lastly, a needle was inserted in the coccygeomesenterial vein and full blood was aspired after which it was centrifuged to collect serum. Intestinal content and bile were freeze-dried and ground fine (<1 mm) for further analysis.

Fig. 1.

Semi-anatomical lay-out of the intestines and accompanying veins in in bird. Sampling points are the duodenum, colon, coccygeomesentarial vein (CMV), peripheral circulation (PC) and the gall bladder (GB).

Quantification of BAs in serum

BAs that were quantified were cholic acid (CA), taurocholic acid (TCA), chenodeoxycholic acid (CDCA), taurochenodeoxycholic acid (TCDCA), ursodeoxycholic acid (UDCA), litocholic acid (LCA), deoxycholic acid (DCA), ursodeoxycholyltaurine (TUDCA), lithocholyltaurine (TLCA) and deoxycholyltaurine (TDA). A 25 µl sample was spiked with 40 µl internal standard (Cholic acid −2,2,4,4-d4, 100 ng/µl, Merck, Darmstadt, Germany). Next, 100 µl ice-cold acetonitrile was added. Subsequently, the samples were vortexed (30 s) and shaken for 10 minutes (175 rpm). Samples were centrifuged (10 minutes, 12 000 x g, 4°C) and supernatants were recovered. To the supernatants, 25 µl methanol and 60 µl ultrapure water were added. Samples were then vortexed (30 s) and filtered (0.22 µm PVDF centrifugal filter). The filtrate was transferred to an HLPC vial with insert to inject in the LC—HRMS system. The chromatographic separation was completed using an Acquity UHPLC with a BEH C18 column (150 mm x 2.1 mm i.d., 1.7 µm; Waters NV, Antwerp, Belgium) with 3 mobile phases (A, B and C). Phase A constituted of ultrapure water with 20 mM ammonium acetate and 1% formic acid; phase B was acetonitrile (ACN) and 1% formic acid and phase C was methanol. The elution program (T0-T10: 25%A-75%C; T10-T13: 15%A-10%B-75%C; T13-T20: 0%A-25%B-75%C) used a flowrate of 0.3 ml/minute. Samples were analysed with a high definition mass spectrometer (Synapt G2 S, Waters NV, Antwerp, Belgium) with negative electrospray ionization (ESI) in MS^E mode. Quantification of the samples was done with a calibration curve and the TargetLynx™ Application Manager (Waters NV, Antwerp, Belgium).

Quantification of BAs in intestinal content

BAs quantified in the intestinal content were cholic acid (CA), taurocholic acid (TCA), chenodeoxycholic acid (CDCA) and taurochenodeoxycholic acid (TCDCA). A 5 mg freeze-dried sample was spiked with 20 µl internal standard (Cholic acid −2,2,4,4-d4, 100 ng/µl, Merck, Darmstadt, Germany), and 105 µl ultrapure water and 125 µl methanol were added. The samples were vortexed (30 s) and shaken (10 minutes, 1400 rpm). Next, samples were centrifuged (10 minutes, 12 000 x g, 4°C) and a 5 µl aliquot of the supernatant was added to 2459 µl ultrapure water and 36 µl internal standard (cholic acid −2,2,4,4-d4, 10 ng/µl). The samples were then vortexed and filtered (Millex Syringe Filter, Durapore® (PVDF) 0.22 µm, Merck Millipore, Darmstadt, Germany). After, the filtrate was transferred to an HLPC vial to inject in the LC—HRMS system. Chromatic separation and quantification was done as mentioned above.

Quantification of BAs in bile

The same BAs that were quantified in serum were also quantified in bile. A 5 mg freeze-dried sample was dissolved in 125 µl ultrapure water and 125 µl methanol. The samples were vortexed (30 s) and shaken (10 minutes, 1400 rpm). Then they were centrifuged (10 minutes, 12 000 x g, 4°C). A 5 µl aliquot of the supernatant was transferred to 2495 µl ultrapure water and vortexed (30 s). Then a 12.5 µl aliquot was added with 40 µl internal standard (cholic acid −2,2,4,4-d4, 10 ng/µl, Merck, Darmstadt, Germany) and 2447.5 µl ultrapure water. The samples were vortexed and filtered (0.22 µm filter). The filtrate was finally transferred to an HLPC vial to inject in the LC—HRMS system. Chromatic separation and quantification was done as mentioned above.

Absorption coefficient of BA

The analysis of TiO₂ as indigestible marker in the freeze dried intestinal content was performed according to Short et al. (1996). Intestinal content of each individual chicken was pooled to obtain enough matrix for analysis of the marker and calculations were performed according to equation (1).

$$absorptioncoeficient=1-\left(\left(\frac{{\left[Ti{O}_2\right]}_{duodenum}}{{\left[Ti{0}_2\right]}_{colon}}\right)*\left(\frac{{\left[BA\right]}_{colon}}{{\left[BA\right]}_{duodenum}}\right)\right)$$ (1)

Data analysis

Statistical analysis was performed in R Studio for Windows (version 1.1.456). The digestibility unit or pen was considered as the experimental unit with animals nested in the pen, and data was analyzed using least-squares linear regression with feed type as independent variable and a post-hoc pairwise comparison was used to assess feed effects. All data were checked for outliers and checked for a normal distribution of the residuals. Differences were considered statistically significant at p < 0.05. Pearsons’s correlation was used to test correlations between DCCF, total bile salt concentration on dry matter, fresh matter and daily BAs excreted.

Results

Animal trial 1

Broilers fed the CS diet showed a higher fat digestibility (p < 0.001) and a trend towards a higher crude protein digestibility than the WRP-diet. Plasma NEFA and TG were also significantly higher in the CS-fed broilers (p < 0.001) (Table 2). The concentration of total BAs in excreta of this group was significantly higher than on the WRP-diet (p < 0.001). This difference in concentration was caused by CDCA, the main component of the bile acid pool (p < 0.001). The concentration of CA was, in contrast, higher in the WRP-fed broilers compared to the CS-group (p = 0.002). Nevertheless, total concentration and the conjugation ratio on fresh matter basis was not significantly different between diets (resp. p = 0.221, p = 0.646). The total BAs lost daily tended to be higher in WRP-fed broilers (p = 0.058).

Table 2.

Effect of diet type (corn-soy based diet (CS) or wheat-palm oil-rapeseed based diet (WRP)) on fecal digestibility coefficients (%) of crude fat (aDCCF), crude protein corrected for uric acid excretion (aDCCP-UA), plasma non-esterified fatty acids (NEFA, mg/dL), plasma triglycerides (TG, mg/dL), total bile salt content (mg/g or µg/g DM and FM), relative conjugated BAs (%) an total daily lost (mg) in excreta of male finisher broilers (25 days) and laying hens (32 weeks) (broiler n = 18; layer n = 12).

	broiler		P	layer		P
	CS	WRP		CS	WRP
aDCCF	84.11a±3.42	54.66b±7.54	<0.001	85.40±2.86	83.05±4.48	0.130
aDCCP-UA	79.51±1.78	78.58±1.79	0.089	65.44±3.61	63.45±3.83	0.211
DM	23.23±3.82	28.37±3.71	<0.001	30.88±2.74	32.16±2.72	0.271
NEFA	0.78a±0.18	0.38b±0.09	<0.001	0.26±0.08	0.29±0.09	0.328
TG	167a±28	104b±35	<0.001	1817±846	1761±402	0.842
BAs
Total (mg/gDM)	1.70a±0.40	1.35b±0.20	<0.001	1.09±0.34	1.35±0.65	0.200
CA (µg/gDM)	66.51±22.45	92.23±30.85	0.002	18.72±34.35	45.52±55.00	0.134
CDCA (mg/gDM)
TCDCA	1.60±0.38	1.21±0.18	<0.001	0.96±0.33	1.14±0.60	0.351
TCA (µg/gDM)	1.46±2.9	2.72±3.71	0.237	0.36±1.25	7.22±2.62	<0.001
TCDCA (µg/gDM)	33.08±24.32	41.67±36.84	0.408	28.75±13.29	122.38±66.73	<0.001
Conjugated (%)	2.03±1.43	2.61±1.42	0.221	3.56b ±3.31	9.95a±4.21	<0.001
Total (µg/gFM)	398.66±122.60	383.02±75.74	0.646	323.03±101.94	437.23±230.08	0.117
Total lost daily (mg)	45.71±12.73	54.04±13.60	0.058	30.83±10.46	41.70±24.90	0.174

Data represents arithmetic means ± 1SD

^a,b superscripts represent significant pairwise post-hoc comparisons (p < 0.05).

DM: dry matter; FM: fresh matter.

In contrast to broilers, the two diet types in laying hens caused no changes in digestibility of fat and protein (resp. p = 0.130, p = 0.211) or the plasma NEFA and TG (resp. p = 0.328, p = 0.842). The total concentration of BAs on dry matter, fresh matter and daily excreted were not significantly different (resp. p = 0.200, p = 0.117, p = 0.174) but both TCA and TCDCA were higher in the WRP-fed layers causing the higher conjugation ratio (p < 0.001).

The total daily excreted BAs weakly but significantly correlated with DCCF (broiler: p = 0.008, r²=0.19; layer: p = 0.016, r²=0.26; data not shown). Similarly total concentration on dry matter (DM) basis yielded significant correlations with DCCF for broilers but not for layers (broiler: p = 0.001, r²=0.27; layer: p = 0.178, r²=0.08). Bile concentration on fresh matter basis did not correlate with fat digestibility (broiler: p = 0.456, r²=0.016; layer: p = 0.212, r²=0.08).

Animal trial 2

The profile of BAs was mainly dominated by the presence of (T)CDCA and (T)CA in all matrices. Half of the bile in the gall bladder dry matter consisted of conjugated BAs of which 86.4% was TCDCA and 12.8% TCA. Duodenal content had a bile salt concentration of 19.6 mg/g DM content, and similarly to the bile, the main BAs found were TCDCA (75.4%) and TCA (11.2%). In addition, the respective deconjugated BAs were identified: CDCA (8.83%) and CA (4.54%). In the colon the relative abundance of unconjugated BAs was highest, but with a total bile salt concentration 10 times lower (1.72 mg/g) than in the duodenum. In the coccygeomesentarial vein the total concentration of BAs was 12 ± 9 µg/ml serum with TCDCA (76.30%), CDCA (4.17%) and TCA (17.74%) being the most abundant bile acids. In the peripheral circulation the bile salt concentration was 6 ± 3 µg/ml serum and, similar to the central vein system, TCDCA (74.66%), CDCA (7.71%) and TCA (14.86%) were the main BAs found. Minor percentages of CA, TCLA and UDCA and DCA were found in both blood pools. Additionally, TUDCA was found in the peripheral circulation blood.

Diet type did not affect the concentration of BAs in the duodenum (p = 0.822) and colon (p = 0.457) (Table 3). The conjugation ratio tended to be lower in the duodenum of Wheat-fed broilers (p = 0.088) and higher in the colon of those broilers (p < 0.001).

Table 3.

Effect of diet type (wheat-corn-soy based diet (Control) or wheat-palm oil-rapeseed based diet (Wheat)) on total bile salt content and conjugated BA ratios in intestines (mg/gDM), blood (µM) or bile (mg/gDM) of male finisher broilers (day 39) (n = 12).

	Control	Wheat	P-value
Intestines
Duodenum
Total (mg/gDM)	19.42±9.40	19.80±5.13	0.822
Conjugated (mg/ gDM)	17.40±8.36	16.78±5.77	0.893
Deconjugated (mg/ gDM)	2.02±1.46	3.02±1.86	0.149
Conjugated (%)	90.09±4.76	83.45±11.10	0.088
Colon
Total (mg/ gDM)	1.63±0.94	1.80±0.39	0.457
Conjugated (mg/ gDM)	0.20b±0.11	0.47a±0.24	0.005
Deconjugated (mg/ gDM)	1.43±0.86	1.33±0.29	0.846
Conjugated (%)	12.73b±4.27	25.70a±9.22	0.001
Blood
mesenterial
Total (µM)	25.56±13.32	20.73±22.15	0.604
Conjugated (µM)	24.12±12.81	19.55±21.71	0.613
Deconjugated (µM)	1.45±0.77	1.17±0.76	0.537
Conjugated (%)	93.98±2.65	90.01±11.43	0.215
Peripheral
Total (µM)	11.23±5.73	12.03±5.03	0.568
Conjugated (µM)	9.46±11.32	11.32±4.84	0.312
Deconjugated (µM)	1.77±1.55	0.71±0.69	0.058
Conjugated (%)	85.85b±7.68	93.86a±4.79	0.009
Bilex
Total (mg/gDM)	508.11±44.78	538.80±26.46	0.073

xAll BAs in the bile were conjugated.

Data represents arithmetic means ± 1SD.

Similarly to the intestine, diet type did not cause differences in concentration in both blood pools (mesenterial: p = 0.604; peripheral: p = 0.568). The ratio of conjugated BAs was not affected in the mesenterial pool (p = 0.215) but a tendency (p = 0.058) for a lower amount of deconjugated BAs was found in the peripheral pool of Wheat-fed broilers resulting in a higher conjugation ratio (p = 0.009). Bile in the gall bladder tended to be more concentrated in Wheat-fed broilers (p = 0.073).

The total bile salt concentration in the peripheral circulation was lower than in the mesenteric veins, with a mean peripheral-to-mesenteric ratio of 0.56 (data not shown). CDCA showed a mean peripheral-to-mesenteric ratio of 0.66 and for the remaining BAs: CA, TCDCA, and TCA, mean ratios were 0.54, 0.51, and 0.44, respectively. In both diets there was a high reabsorption degree for conjugated BA, despite a tendency that Control-fed broilers tended to reabsorb TCA and TDCA more (resp. p = 0.055, p = 0.054) (Table 4). For the unconjugated BA, CA showed a higher degree of reabsorption than CDCA but both were not affected by diet (resp. p = 0.504, p = 0.781).

Table 4.

Effect of diet type (corn-soy based diet (Corn) or wheat-palm oil-rapeseed based diet (Wheat)) and effect of conjugation on total bile salt digestibility coefficient (%) between duodenum and colon of male finisher broilers (day 39) (n = 12).

	Control	Wheat	P-value
TCA	99.86±0.14	99.66±0.25	0.055
TCDCA	99.61±0.34	99.25±0.47	0.054
CA	96.15±4.81	97.49±3.02	0.504
CDCA	78.47±20.06	75.41±20.49	0.781
total	97.55±2.43	97.81±0.62	0.811

Data represents arithmetic means ± 1SD.

Discussion

The bile salt profile in the various matrices was mainly limited to the presence of (T)CDCA and (T)CA, in agreement with the study of Elkin et al. (1990). Nevertheless, there was one study that found considerable amounts of lithocholylglycine in chickens (Yeh and Hwang, 2001). There is a general consensus that most bird species do not conjugate BAs with glycine, although scarce exceptions in low quantities exist (Hagey et al., 1994, Hofmann et al., 2010; Tancharoenrat et al., 2022). Here, we focussed on the taurine conjugated BAs which may imply an underestimation of total BAs if glycine conjugation would exist in chickens. Next to the main BAs, (T)CDCA and (T)CA, there were traces of (T)LCA, DCA and (T)UDCA in the bile and blood, indicating modest modification by the microbiome into secondary BAs, or secondary BAs showed a low degree of intestinal absorption. Due to its low abundance we did not further analyse these minor BAs in the excreta, therefore the true pathway of the secondary BAs is still elusive and subject to further research.

Our intended diet challenge resulted in the anticipated effect, i.e. a distinct difference in fat digestibility and a lower plasma triglyceride concentration as indicator of reduced fat absorption. The lower circulating NEFA in the WRP-fed chickens typically reflects fat status in response to a lower energy balance. The mild reduced crude protein digestion in relation to the large effect of lower fat digestion could result in lower energy:protein ratios in WRP-diet (Mansilla et al., 2022). This causes growing broilers to have lower fat reserves hence lower NEFA mobilisation. The impact of this diet on digestibility may relate to the dietary fatty acid profile: soy oil is mainly composed of C18:2, while palm oil has C16:0 as the main fatty acid; leading to an unsaturated/saturated ratio of 4-5.5 and 1, respectively (Ravindran et al., 2016; Wealleans et al., 2021). It is known that saturated fats are harder to emulsify and digest, and hence may require more BAs (Tancharoenrat et al., 2014; Ravindran et al., 2016). However, in this study excreta bile acid concentration on DM basis decreased on the W-diet which may originate from the difference in DM content of the intestinal content. On fresh matter basis there is no difference in abundance of BAs in the excreta between diets. Moreover, daily excreted BAs tended to be higher in the WRP-fed animals. Daily bile acid excretion may serve as a potential indicator of fat digestibility. However, the observed correlation was weak and likely induced by increased total excreta output associated with higher dry matter content. Using bile acid concentration to assess and interpret fat digestibility appears to have limited accuracy and applicability across both poultry genotypes.

The strong influence of diet on broiler metabolism made them ideal models for the study of bile salt dynamics in galliform birds. In the second trial, the Wheat-based broiler diet tended to generate more BAs in the gall bladder, and could be linked to a higher need for BAs to emulsify the saturated palm fat in this diet and be a result of the efficient reabsorption of bile (Ravindran et al., 2016).

Despite the large difference in fat digestibility in the first trial, BA concentrations were similar between diets in both duodenum and colon on the second trial. However, more conjugated BAs were found in the colon leading to the higher conjugation ratio in the Wheat-group. This finding may align with the numerically higher conjugation ratio in the excreta of Wheat-fed birds in trial 1. The bile salt conjugation ratio would not directly depend on feed composition but on microbial composition (Foley et al., 2019). Some bacteria residing in the poultry gut, such as Enterococcus faecium, Clostridium perfringens and Lactobacillus spp., produce bile salt hydrolases (Knarreborg et al., 2002; Moser and Savage, 2001), interfering with the deconjugation ratio. This could be linked to the effects of antimicrobial drugs on digestibility of nutrients (Guban et al., 2006). We did not assess the intestinal microbiome profile, but it is well documented that the intestinal microbial communities on a wheat-based diet are distinctly different from those on a corn-based diet (Teirlynck et al., 2009). Although comparable in BA concentration, colon content and excreta are two different matrices and the extent of comparison might a matter of debate due to the post-excretory microbial modification.

The intestinal recovery rate in the colon versus duodenum tended to be higher in control-fed birds however the difference might be of biological insignificance. Moreover, the more concentrated bile salt in bile of the Wheat-fed birds may indicate that more BAs are in circulation which makes the enterohepatic cycle more prone to loss in the excreta. This was also seen in the higher total BAs daily excreted in trial 1. The uptake of conjugated BAs is regulated by the apical sodium-dependent bile salt transporter (ASBT) mainly found in the ileum, whereas unconjugated BAs can cross passively through the gut barrier (Hofmann and Hagey, 2008; Nakao et al., 2015). The high absorption ratio of conjugated BAs versus the significantly lower ratio of deconjugated BAs in this study could demonstrate this. However, to a lesser extent BAs could already be absorbed from the duodenum (Webling, 1966). This might be indicated by the decreased conjugation ratio found in the duodenum due to the active reabsorption of the conjugated BAs. A downregulation of the ileal bile salt transporters to maximize intestinal concentration on conjugated BAs and fat emulsification could be suggested to occur in wheat-fed broilers, since saturated fatty acids are digested and absorbed further along the intestinal tract compared to polyunsaturated fatty acids (Tancharoenrat et al., 2014).

In mammals, BAs are absorbed in the hindgut, transferred via the mesentery veins towards the liver where they are filtered and stored in the bile bladder, known as the enterohepatic loop (Redinger, 2003; Hofmann and Hagey, 2008). This cycle would result in high BA concentrations in the mesentery veins and low concentrations in peripheral circulation. Based on the limited availability of reference values in birds, bile salt concentrations are generally high in peripheral circulation in reference to healthy humans which of which the most abundant BAs (GCDCA, CDCA and DCA) have a range of 0.03 to 2 µM (Joseph et al., 2025). Of the following phylogenetic orders following ranges of total BAs were reported Columbiformes (n = 2) [22-60 µM], Gallonserae (n = 6) [36-74 µM], Passeriformes (n = 1) [23-90 µM], Psittaciformes (n = 26) [23-104 µM], Strigiformes (n = 8) [7-66 µM] (Green and Kellogg, 1987; Harr, 2002; Cray and Andreopoulos, 2003; Carpenter and Marion, 2017). This study demonstrated that peripheral bile salt levels were lower than mesenteric levels, however with a mean peripheral:portal ratio of 56%, they are still relatively high in relation to portal levels. Few studies in literature have compared BA concentrations between the portal and systemic circulation. In fasted humans, systemic concentrations of cholic acid (CA), chenodeoxycholic acid (CDCA), taurocholic acid (TCA), and taurochenodeoxycholic acid (TCDCA) were reported to be approximately 8%, 18.4%, 14%, and 26%, respectively, of portal vein concentrations (Ahlberg et al., 1977; Linnet et al., 1984). Van den Berghe-Henehouwen and Hofmann (1983) showed that dihydroxy BAs, particularly (T)CDCA, are taken up less efficiently by the liver than (T)CA, resulting in higher systemic concentrations and increased dihydroxy:trihydroxy ratios. In addition, conjugated BAs are generally considered to be taken up more efficiently by the liver than their unconjugated counterparts. In the present study, however, chickens were in a fed state, which may have influenced the observed systemic:portal ratios. Angelin et al. (1982) demonstrated that, in humans, the peripheral:portal ratios of CA (9.95%) and CDCA (21.26%) in the fasted state decreased postprandially to 8.41% and 15.75%, respectively. This decrease was attributed to an increase in portal concentrations, while peripheral concentrations remained relatively stable, a phenomenon also observed in pigs (Eggink et al., 2018). Accordingly, the relatively high ratios observed in fed chickens in our study would be expected to be even higher under fasting conditions.

Despite differences in systemic:portal ratio, the profile of chicken BAs were largely similar between the two circulations. Under normal conditions in mammals, unconjugated BAs are thought to escape hepatic uptake more readily, whereas conjugated BAs are taken up more efficiently by the liver (van den Berghe-Henehouwen and Hofmann, 1983). In contrast, our findings in chickens do not fully support this concept, as conjugated BAs do not appear to be preferentially extracted and showing lower quantities in peripheral circulation. Notably, the high levels of TCDCA detected in the systemic circulation suggest either an altered hepatic uptake capacity for this bile acid or a partial bypass of the liver, indicating potential species-specific differences in hepatic bile acid handling.

We hypothesize that birds due to their renal portal system could shunt the liver via the kidney directly into general circulation (Buyse et al., 2025). In birds, the CMV is directly linked to the hind part of the kidney portal system and, via the renal portal valve, blood could be injected in the general circulation, thus bypassing the liver. Moreover, blood in the CMV could flow in both directions (Akester, 1964), as indicated by the arrow in Fig.1. The higher presence of deconjugated BAs found in the peripheral circulation would enforce the hypothesis of blood bypassing the liver where amino acid N-acyltransferase conjugates the BAs with taurine, hence conjugation ratio should be higher. Moreover, the conjugation ratio in periphery is higher in the Wheat-fed broilers which may relate to transformations in the colon. Since reabsorption from the ileum is closer to the CMV we can assume that BAs resorbed in the hindgut are more prone to be circulated towards the periphery. This effect suggests that the enterohepatic cycle could be redefined in birds and should be subject for further research.

Conclusion

BA metabolism in birds may be more complex than previously thought. In growing broilers, their concentration and composition can be modulated by diet but also the microbiome. The bile salt profile in broiler chickens is limited to basically (T)CDCA and (T)CA, showing that the microbiome is more focused on deconjugating BAs rather than converting them to secondary BAs. Based on the absorption ratio, the efficiency of the enterohepatic cycle is comparable to mammals. However, the higher concentration ratio of BAs in the general circulation compared to the mesenteric circulation, together with the unique anatomical layout of birds, suggests the possibility to redefine the avian enterohepatic circulation.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Kobe Buyse: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Christof Van Poucke: Writing – review & editing, Methodology, Investigation. Marta Lourenço: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization. Geert PJ Janssens: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Disclosures

None of the authors has any financial or personal relationships that could inappropriately influence or bias the content of the paper.