Detection of Bifidobacterium animalis subsp. lactis (Bb12) in the Intestine after Feeding of Sows and Their Piglets

Abstract

A real-time PCR method has been developed to distinguish Bifidobacterium animalis subspecies in the gastrointestinal tracts of pigs. Identification of a highly conserved single-copy tuf gene encoding the elongation factor Tu involved in bacterial protein biosynthesis was used as a marker to differentiate homologous Bifidobacterium animalis subsp. lactis (strain Bb12) from Bifidobacterium animalis subsp. animalis, as well as Bifidobacterium suis, Bifidobacterium breve, Bifidobacterium longum, several species of Lactobacillus, and Enterococcus faecium. Real-time PCR detection of serially diluted DNA extracted from a pure culture of Bb12 was linear for bacterial numbers ranging from 10 to 10,000 tuf gene copies per PCR (r² = 0.99). Relative differences in Bb12 bacterial numbers in pigs fed daily with Bb12 were determined after detection of Bb12 tuf gene copies in DNA extracted from the intestinal contents. Piglets treated with Bb12 immediately after birth maintained a high level of Bb12 in their large intestines with continuous daily administration of Bb12. Piglets born to Bb12-treated sows during the last third of their gestation and also treated with Bb12 at birth (T/T group) had a higher number of Bb12 organisms per gram of intestinal contents compared to placebo-treated piglets born to placebo-treated sows (C/C group), Bb12-treated sows (T/C group), or piglets born to placebo sows but treated with Bb12 immediately after birth (C/T group). In addition, there was a significant increase in gene expression for Toll-like receptor 9 (TLR9) in piglets from the T/T group, with no change in TLR2 and TLR4. These findings suggest that the tuf gene represents a specific and functional marker for detecting Bifidobacterium animalis subsp. lactis strain Bb12 within the microbiota of the intestine.

Bifidobacteria are anaerobic, gram-positive, non-spore-forming, non-motile bacilli commonly found in the gastrointestinal tracts (GITs) of animals, including humans (1). Bifidobacteria are the predominant bacterial species in the GITs of infants; they represent about 3% of the total microbiota in the intestine of healthy adult humans (16) and are associated with beneficial health effects (15, 16, 30, 32, 36). Despite the general acceptance of bifidobacteria as a probiotic, and their use in health-promoting foods such as fermented milks, infant formula, cheese, and ice cream, there is little definitive information to support a mechanism of action. Stimulation of host resistance, immune modulation, and competitive exclusion of pathogens, however, have been proposed as likely mechanisms (40).

One of the Bifidobacterium species commonly used in the food industry is Bifidobacterium animalis subsp. lactis strain Bb12, which is marketed around the world under a variety of labels in dairy products and infant formulas (37, 40). The taxonomy of B. animalis subsp. lactis has been controversial since its original description by Meile et al. in 1997 (27), and several studies have investigated its similarity with the closely related species Bifidobacterium animalis subsp. animalis (41). New genotypic evidence reported by Ventura et al. (46-48), Zhu and Dong (54), Masco et al. (23), and Kwon et al. (19) indicate that B. animalis subsp. lactis and B. animalis subsp. animalis should be considered two separate taxonomic entities at the subspecies level. B. animalis subsp. lactis exhibits properties such as elevated oxygen tolerance (34), differential growth in milk-based media (46), and hydrolysis of milk proteins (13); these properties differ from B. animalis subsp. animalis and facilitate its growth in commercial products under nonanaerobic conditions. Traditional bacteriological and biochemical identification techniques, such as selective growth of species in differential media, cannot be routinely used to differentiate Bifidobacterium species. These methods are time consuming and limited by low sensitivity and reproducibility due to the multitudes of species that grow and require further identification (24). In addition, the information obtained by culture-based growth methods provides only a fragmented picture of the relative distribution of species within the GIT because a significant part of its microbiota cannot be grown in vitro (43, 55). Culture-independent methods have been developed in recent years as an alternative to characterize whole bacterial communities by direct extraction of DNA from fecal samples without prior cultivation (6). These methods include fluorescent in situ hybridization, dot blot hybridizations, and DNA arrays and fingerprinting methods such as terminal restriction fragment length polymorphism and denaturating or temperature gradient gel electrophoresis (26). Conserved and variable regions within the 16S ribosomal gene are widely used as markers to study bacterial diversity by PCR with sensitivity that is approximately 100 times greater than that of traditional culture-based and fluorescent in situ hybridization methods (25). The 16S ribosomal gene variable regions may be utilized for genus or species differentiation if species-specific probes can be designed (19, 25, 39). Highly homologous species like B. animalis subsp. lactis, however, are not generally distinguished by 16S ribosomal gene-based differentiation (28, 38, 53) or quantitative assessment. In addition, the use of 16S ribosomal gene-based probes for quantitative real-time PCR remains difficult because the copies of ribosomal DNA per genome can vary (5). An alternative to 16S ribosomal gene-based analysis of Bifidobacterium species is comparing conserved protein coding sequences of bacterial genes, such as those for transaldolase (35), recA (17, 47), hsp60 (14, 54), groEL (49), groES (49), tuf (48), atpD (49), dnak (50), or xfp (52). After comparing the discriminating properties of each of these sequences, we selected the highly conserved and ubiquitous tuf gene encoding the elongation factor Tu that facilitates the elongation of polypeptides from the ribosome and aminoacyl tRNA during translation and that has been used as a phylogenetic marker for eubacteria (31, 33). The tuf gene is universally distributed in Bifidobacterium and Lactobacillus species, and only one tuf gene per bacterial genome has been found (7, 48). It was the only gene that was able to discriminate closely related B. animalis isolates at the subspecies level. The objective of this work was to validate the use of a real-time PCR assay to detect the single-copy tuf gene of B. animalis subsp. lactis as a marker for Bb12 in the GITs of pigs orally treated with Bb12. In addition, the effect of different dietary exposures to Bb12 on the number of organisms detected in the intestinal contents of pigs and on the host innate immune response was assessed by measuring the localized gene expression of Toll-like receptors (TLRs). TLR2, -4, and -9 were chosen because of their abilities to bind to bacterial products that activate the innate immune system and influence the course of acquired immunity (2). We demonstrate, for the first time, the use of the tuf gene to quantitatively detect Bb12 in the GITs of Bb12-treated piglets and the in vivo effect on host TLR expression. There is also a significant maternal effect on Bb12-treated piglets from Bb12-treated sows that results in higher and more persistent levels in the proximal colon. The sensitivity of the molecular assay used to detect B. animalis subsp. lactis should prove useful in supporting health benefits associated with the probiotic efficacy of Bb12.

MATERIALS AND METHODS

Reference strains and culture conditions.

The bacterial strains used to design and validate the quantitative assay for B. animalis subsp. lactis (strain Bb12) are listed in Table 1. All bifidobacteria were cultured on Bifidobacterium-selective medium (Bifido; Anaerobe Systems, Morgan Hill, CA). Lactobacilli were cultured on Lactobacillus-selective MRS agar (Difco Laboratories, Detroit, MI), and enterococci were cultured on Enterococcus-selective agar (Difco Laboratories, Detroit, MI). The plates were incubated at 37°C for 3 days in a Bactron IV anaerobic chamber (Sheldon Manufacturing, Inc., Cornelius, OR). Colony formation was examined with a stereoscopic microscope, and bacterial characteristics were determined by Gram staining. Bacteria were swabbed from the plate, resuspended in 1 ml of sterile phosphate-buffered saline (PBS), and stored at −20°C until required for further processing for DNA extraction. Serial 10-fold dilutions of 1 g of lyophilized B. animalis subsp. lactis strain Bb12 organisms, provided by Chr. Hansen (Milwaukee, WI), were plated on anaerobic Brucella blood agar (BRU; Anaerobe Systems) and Bifido plates to test purity. The plates were subsequently incubated at 37°C for 3 days in an anaerobic chamber with a gas mixture of 5% CO₂, 5% hydrogen, and 90% nitrogen. Bacteria were identified by Gram staining and with an API 20A anaerobe identification strip (bioMerieux, Hazelwood, MO). CFU were determined in duplicate. The highest dilution showing growth of bacteria was used for the final CFU determination.

TABLE 1.

Bacterial strains examined

Species	Straina	Sourceb
Bifidobacterium
B. animalis subsp. lactis	Bb12	Chr. Hansen
	27536	ATCC
B. animalis subsp. animalis	25527T	ATCC
B. suis	27531	ATCC
B. breve	15700T	ATCC
B. longum	15707T	ATCC
	BB46	Chr. Hansen
Other
Enterococcus faecium	SF273	Chr. Hansen
Lactobacillus paracasei	LC-01	Chr. Hansen
Lactobacillus bulgaricus	LBA40	Chr. Hansen
Lactobacillus acidophilus	53544	ATCC
	53545	ATCC
	LA05	Chr. Hansen

^aT, T type strain.

^bChr. Hansen strains were obtained from the Chr. Hansen Collection. All other strains were obtained from ATCC.

Study design and sample collection.

The presence of Bb12 was evaluated in intestinal contents or fecal samples taken from Bb12-treated or placebo-treated pigs from three independent experiments using pregnant sows and their litters. In the first experiment, four pregnant sows were given an oral daily dose of 5 g (3.7 × 10¹⁰ CFU/sow/day) of freeze-dried Bb12 during the last third of their pregnancy. Four additional sows were given an oral daily dose with 5 g of the placebo containing only the vehicle used in the probiotic product. Immediately after birth, piglets born from each sow received either a daily probiotic treatment of 1.5 g (1.05 × 10¹⁰ CFU/day) or an equivalent amount of placebo containing only the vehicle for 32 days. This experimental design gave the following four different experimental groups of pigs: (i) Bb12-treated sows and Bb12-treated piglets (T/T) (n = 18); (ii) Bb12-treated sows and placebo control-treated piglets (T/C) (n = 16); (iii) placebo control-treated sows and Bb12-treated piglets (C/T) (n = 21); and (iv) placebo control-treated sows and placebo control-treated piglets (C/C) (n = 19). An additional fifth litter (n = 9) of untreated piglets from an untreated sow located in a different region of the farrowing barn was used as a negative control for probiotic and placebo treatment. All 83 piglets were weaned at day 21 and euthanized on day 32 after birth. Five grams of intestinal contents from the proximal colon was taken immediately after necropsy and kept frozen at −20°C until further processing for DNA extraction. Both products (probiotic and placebo) were microbiologically tested throughout the experiment for purity and stability of the bacteria in the probiotic product and for the absence of bacterial growth in the placebo product.

In a second replicate experiment, fecal samples were collected from a total of 20 piglets born to sows treated with either Bb12 (3.7 × 10¹⁰ CFU/sow/day) or placebo during the last third of gestation. Piglets also received a daily dose (1.05 × 10¹⁰ CFU/day) of lyophilized Bb12 or the equivalent amount of placebo starting at birth and given until day 23 (the weaning date). This experimental design assigned five piglets per experimental group (T/T, T/C, C/T, and C/C). Fresh fecal samples were aseptically collected in a sterile 50-ml conical tube after pigs were manually stimulated to defecate. At least 1 g of feces was collected at 10, 23, 32, and 45 days of age. Samples were immediately stored at −20°C and kept frozen until DNA extraction.

In a third replicate experiment, three sows were orally inoculated daily during the last third of their pregnancy and through weaning of their piglets at 19 days after birth with 2.2 g of freeze-dried Bb12 (3.52 × 10¹⁰ CFU/sow/day). Three additional sows were inoculated with the same amount of a placebo preparation. Piglets within each litter were randomly divided at birth into two groups, where half of the litter received a daily probiotic treatment of 1.1 g (1.76 × 10¹⁰ CFU/pig/day) and the other half received an equivalent amount of placebo through weaning at 19 days after birth and for an additional 72 days postweaning. This experimental design gave four groups of pigs that included the following: (i) T/T piglets (n = 14); (ii) T/C piglets (n = 17); (iii) C/T piglets (n = 13); and (iv) C/C piglets (n = 13). Five grams of intestinal contents from the proximal colon was collected at necropsy and stored at −20°C until needed for further processing for DNA extraction. Sets of three to five piglets were euthanized at 7, 19, 32, and 91 days after birth, and intestinal samples were collected from the proximal colon along with a 2-cm² tissue section of intestinal mucosa. All animal procedures were approved by the Beltsville Area Animal Care and Use Committee.

DNA extraction.

Reference bacteria cell suspensions (Table 1) were centrifuged for 10 min at 5,000 × g. The bacterial pellet was resuspended and incubated for 30 min at 37°C with enzymatic lysis buffer (20 mM Tris-Cl [pH 8.0], 2 mM sodium EDTA, 1.2% Triton X-100, and 20 mg/ml lysozyme [Sigma, MO]), followed by an incubation with proteinase K and buffer for 30 min at 70°C. After enzymatic lysis was carried out, bacterial DNA was isolated from the samples using the DNeasy tissue kit (Qiagen, Valencia, CA) according to the instructions of the manufacturer. The DNA was eluted in TE buffer (10 mM Tris-HCl, 1 mM EDTA [pH 8.0]).

Similarly, DNA from the fecal contents of animals was isolated using the QIAamp DNA stool mini kit (Qiagen, Valencia, CA). Briefly, 1 g of homogenized contents from different intestinal sites was thawed, weighed, and resuspended with lysis buffer. After heating the suspension at 95°C for increased DNA yield, removal of inhibitors and proteinase K digestion were done before DNA was bound to a column, washed, and eluted in TE buffer. DNA concentration was determined by spectrophotometry. An aliquot of 100 ng of DNA from each extraction was used as a template for all bacterial quantifications.

Primers specific for Bifidobacterium animalis subsp. animalis.

Complete and partial sequences of the 16S to 23S intergenic spacer region of several Bifidobacterium species and Enterococcus faecium were retrieved from GenBank (Table 2) to develop primers and probes for 5′ nuclease assays. All published sequences in GenBank for the dnaK (50), groES, groEL, atpD (49), tuf (47), recA (47), hsp60 (14), and transaldolase (35) genes of Bifidobacterium species were aligned using the Clustal program (44), and the overall nonconserved regions of these sequences were used to design primers and probes for the detection of B. animalis subsp. animalis. To increase the specificity and sensitivity of the assay, TaqMan minor groove binding probes were used. The primers and probes were designed using Primer Express software (Applied Biosystems, CA). The oligonucleotide probes designed were labeled with the 5′ reporter dye 5′-tetrachloro-fluorescein phosphoramidite and the 3′ quencher BHQ1 (Biosource, CA) (Table 3). For determination of the total bacterial load, a previously described set of primers and probes for the 16S rRNA genes of eubacteria labeled with the 5′ reporter dye 6-carboxyfluorescein and the 3′ quencher NFQ-MGB (Applied Biosystems, CA) was used (29). Similarly, the primers and probes used to detect total Bifidobacterium spp. (10) and Lactobacillus spp. (12) were used as described previously. All primers were tested for specificity using the Clustal alignment tool (44) and bacterial reference strains as templates for real-time PCR analysis. The assays were performed with a 25-μl PCR amplification mixture containing 1× Thermo-Start QPCR master mix with ROX (ABgene, Rochester, NY), 50 to 300 nM of forward and reverse primer, 100 to 200 nM of probe, and an equivalent of 100 ng of bacterial DNA. The bacterial DNA concentration was determined by spectrophotometer (Beckman Coulter DU640, Fullerton, CA). The amplification conditions were 50°C for 2 min, 95°C for 10 min, and 40 cycles at 95°C for 15 s and 60°C for 1 min. Fluorescent signals measured during amplification were processed postamplification and were considered positive if the fluorescence intensity was >20-fold of the standard deviation of the baseline fluorescence. This level was defined as the threshold cycle (C_T) value and is inversely correlated to the amount of nucleic acid in the original sample (C_T value of 40 = no amplification). C_T values for each assay were compared among reference strains to establish discriminatory properties between homologous Bifidobacterium species.

TABLE 2.

GenBank sequence used to design real-time PCR assays

Species	Strain	GenBank accession no. for:
		groES	hsp60, groEL	recA	16S to 23S	Transaldolase	atpD	grpE	dnaK	tuf
Bifidobacterium	ATCC 15703						AY487144			AY372045
adolescentis	CIP6459				U09512
	CIP6460				U09513
	CIP6461				U09514
	DSM 20083					AF417530
	JCM 1275	AY585248						AY642885	AY642864
Bifidobacterium	ATCC 27535					AF417536
angulatum	JCM 7096	AY585256						AY642880	AY642870
Bifidobacterium	LMG 11615							AY642876
animalis	ATCC 25527	AY585250	AY488178	AY370929	U09858		AY487152	AY642878	AY642871	AY370920
	ATCC 27536		AY488181	AY370923	L36967		AY491983			AY370912
	ATCC 27672		AY488183	AY372028			AY491986			AY370921
	ATCC 27673			AY372029						AY370917
	ATCC 27674		AY488179	AY372030			AY491988			AY370918
	LMG 11615							AY642887	AY642876
	LMG 18900							AY642890	AY642872
	LMG 18906							AY642889	AY642873
	NCC 239		AY488182	AY370924			AY491984			AY370913
	NCC 311			AY370925						AY370914
	NCC 330						AY491985
	NCC 363		AY488176	AY370928						AY370915
	NCC 383			AY370926						AY370922
	NCC 402		AY488177	AY370927			AY491987			AY370916
Bifidobacterium animalis subsp. lactis	DSM 10140		AY488180	AY372031	X89513	AF417537	AY487153	AY667066	AY642912	AY370919
Bifidobacterium	ATCC 15696			U50267
bifidum	ATCC 29521						AY487145			AY372041
	CIP567				U09517
	CTP6465				U09831
	DSM 20456					AF417538
	DSM 20456					AF417533
	JCM 1255	AY585252						AY642888	AY642868
Bifidobacterium	ATCC 15698				U09518	AF417532
breve	ATCC 15700						AY487154			AY372046
	ATCC 15701			U50268
	CIP6468				U09519
	CIP6469				U09520
	CIP6470				U09521
	NCFB 2258			AF094756
	UCC2003	AY585262	AY585261
	Y8				AJ245850
Bifidobacterium	ATCC 27539				U09522		AY487146
catenulatum	DSM 20103					AF417534				AY372044
	JCM 1194	AY585249						AY642884	AY642875
Bifidobacterium choerinum	ATCC 25911						AY487148
Bifidobacterium	ATCC 25911				U09523
coryneforme	ATCC 27686						AY487147
	JCM 5819	AY585258
Bifidobacterium cuniculi	ATCC 27916				U09790
Bifidobacterium	ATCC 27534				U10434		AY487149
dentium	JCM 1195	AY585247						AY642886	AY642866
Bifidobacterium globosum	ATCC 25865				U09524
Bifidobacterium indicum	ATCC 25912				U09791
Bifidobacterium	ATCC 15697			U50269	U09792	AF417529	AY487150
infantis	ATCC 15702					AF417540
	ATCC 25962				U09525
	CIP6378				U09527
	JCM 1222	AY585254						AY642882	AY642867
	Y1				AJ245851
Bifidobacterium animalis subsp. lactis	LMG 18906	AY586538	AY586539
Bifidobacterium	ATCC 15707		AY835622			AF417531				AY372043
longum	ATCC 15708			U50270	U09832
	ATCC 27533						AY487151		AY642869
	NCC2705	NC_004307	NC_004307	NC_004307	NC_004307	NC_004307	NC_004307	NC_004307	NC_004307	NC_004307
	Y10				AJ245849
Bifidobacterium longum subsp. suis	JCM 1269	AY585253						AY642883
Bifidobacterium	ATCC 27540				U09878
magnum	JCM 1218	AY585251						AY642877	AY642863
Bifidobacterium	ATCC 27919					AF417535
pseudocatenulatum	JCM 1200							AY642881	AY642865
Bifidobacterium	ATCC 25526				U09879
pseudolongum	JCM 5820	AY585260
Bifidobacterium pullorum	JCM 1214	AY585255
Bifidobacterium	ATCC 25525				U09528
thermophilum	JCM 1207	AY585257						AY642879	AY642874
Enterococcus faecium	ATCC 19434		AF417582

TABLE 3.

Primers and probes used in the duplex 5′ nuclease assays

Target gene	Primer/probe	Sequence (5′ → 3′)	Amplicon length (bp)
Transaldolase gene	Forward	CGA CAA GAA GCT CGA GGA GAT	116
	Reverse	CGG ATC CTC GGC GAA CT
	Probe	CCT TGC CTT CGA GAC CCT TGG CCT
Transaldolase group gene	Forward	GCG TCC GCT GTG GGC	106
	Reverse	CTT CTC CGG CAT GGT GTT
	Probe	TCC ACC GGC ACC AAG AAC GC
atpD	Forward	GAT GTT ACC AAG GGC CAT GTG	83
	Reverse	CGC TCC TTG ATC ACG ATC TTC T
	Probe	CGA CGT TTC CGG CCA CAT TCT CA
	BGB probe	TTC CGG CCA CAT TC
dnaK	Forward	GCA GCT CTG GCC TAC GGT	189
	Reverse	ATA ATG CGC TGG TCC CAA TC
	Probe	CCC GAC GTA GCC TGC ACC TGG
	BGB probe	AGG CTA CGT CGG GCG
groES	Forward	TTG GCC CAG GTC GTC GT	119
	Reverse	AGG TAT TCC TCG CCC TTG AAG T
	Probe	AGG GCG AGC GTG TTC CCA TGG A
	BGB probe	CGT GTT CCC ATG GAC
16S to 23S	Forward	TTT GCC GAG TGC GAT GGT	104
	Reverse	GTG GCG GCC AGG GAA C
	Probe	CCT GGC TTG CTG GCG TGG AAG AG
groEL, hsp60	Forward	CCA AGT GGG TAA GCA TGA ATT TC	109
	Reverse	GGT ATC GGC CAG CTT ATC CA
	Probe	CCT GAC GAG CTT CCT CAT CGT ATT CAA TG
	BGB probe	CAA AGA TCA TTG AAT ACG ATG AG
recA	Forward	GAA GGC GAT ATG GGT GAC AG	133
	Reverse	GCC GAT CTT CTC TCG CAA CT
	BGB probe	CAC AGG CGA ACA CGA
tuf	Forward	GTG TCG AGC GCG GCA A	117
	Reverse	CTC GCA CTC ATC CAT CTG CTT
	BGB probe	ATC AAC ACG AAC GTC GAG A
B. breve groEL, hsp60	Forward	ATG TTG ACG GCG AGG CTC	85
	Reverse	AAC CCG GGT GCT TTG ACA
	Probe	CCC TGA TTC TGA ACA ACA T
E. faecium groEL, hsp60	Forward	GAA ACG ACG GTG TCA TCA	126
	Reverse	TCC ATT TTG TCG TTA TCT G
	Probe	CCA CGA TCA AAC TGC ATA C

Specific primers for Bifidobacterium animalis subsp. lactis (strain Bb12).

The amount of Bb12 in culture or in intestinal samples was determined using the primer-probe set for the tuf gene (Table 3). A genomic tuf gene fraction was amplified and used as an internal control and to generate a standard curve. The size of the fragment (117 bp) and its molecular mass were quantitatively determined on the DNAChip (Agilent, Waldbronn, Germany) using Bioanalyzer (Agilent Technologies, Wilmington, DE). The molecular mass was used to calculate the copy number of the target amplicon used to generate a series of six serial dilutions of known concentrations using a range that matched the expected concentration range of the unknown samples. Serial dilutions were analyzed in triplicate by real-time PCR in separate sample wells but within the same run, and the resulting C_T values were recorded. The unknown samples were evaluated within the testing interval known to be linear. A plot of C_T versus the logarithm of the copy number corresponding to that C_T resulted in a straight line standard curve. The number of target gene copies was then extrapolated from the standard curve equation. All dilutions of unknown samples were run in triplicate, and samples were diluted in TE buffer using salmon sperm as a carrier (Invitrogen, CA) at a concentration of 60 μg/ml.

Sensitivity of the tuf gene assay for detection of Bb12.

The sensitivity of the tuf gene duplex 5′ nuclease assay was assessed by testing for a series of diluted samples of exogenous Bb12 added to fecal samples of control pigs (those not exposed to probiotic treatment). One gram of lyophilized Bb12 was cultured on MRS agar and incubated anaerobically for 3 days at 37°C to verify purity and for quantitative determination of CFU per gram. Ten different 1-gram aliquots of homogenized feces of pigs not previously exposed to the probiotic were spiked with either nine 10-fold dilutions of 1 gram of lyophilized Bb12 previously dissolved in PBS or with 10-fold dilutions of a pure Bb12 broth culture. After total DNA extraction from spiked feces, serial 10-fold dilutions were analyzed in triplicate by real-time PCR in separate sample wells but within the same run, and the resulting C_T was compared to the standard curve analysis to estimate the bacterial copy number in each serial dilution. The efficiency of the PCR amplification for the Bb12 tuf assay was calculated for different matrices (PBS and feces) using the following formula as stated by Bustin and Nolan (4): efficiency = 10^{(−1/slope)−1}. Similarly, samples from intestinal contents of pigs that received Bb12 orally for different periods of time were used to evaluate the detection level of the Bb12 tuf gene assay.

Host mRNA gene expression by quantitative real-time PCR.

The mRNA levels of the TLR2, TLR4, and TLR9 genes were detected in the proximal colon of piglets derived from experiment 3. Tissue sections (2 cm²) were dissected from the proximal colon mucosa of pigs from the four treatment groups at days 7, 19, 32, and 91 and were immediately frozen in liquid nitrogen and stored at −70°C. Tissue RNA was extracted after homogenization in Trizol reagent (Invitrogen, Gaithersburg, MD), and quantitative real-time PCR was performed on cDNA synthesized from each sample using 10 μg of total RNA (42). RNA was treated with DNase in the presence of RNA inhibitor (Ambion, CA). The absence of genomic DNA contamination was confirmed after no signal was detected when running a PCR using a non-exon-spanning probe for the RPL32 housekeeping gene. DNase-treated RNA was quantified using Bioanalyzer 2100 and the RNA 6000 Labchip kit (Agilent Technologies, Palo Alto, CA) (9). Briefly, cDNA was synthesized with Superscript RT (Invitrogen), and oligo(dT) and 50 ng of this cDNA were used for real-time PCR amplifications using a Thermo-Start DNA polymerase master mix (ABgene, Rochester, NY) and the ABI Prism 7700 sequence detector system (Applied Biosystems, Foster City, CA). Amplification conditions were as follows: 50°C for 2 min; 95°C for 10 min; 40 cycles of 95°C for 15 s; and 60°C for 1 min. All probes and primers selected for real-time PCR were designed using the Primer Express software package (Applied Biosystems, Foster City, CA), and nucleotide sequences were obtained from GenBank or the TIGR porcine EST database. The TLR2, TLR4, and TLR9 genes were selected to evaluate innate immune response activation by bacterial ligand. The sequence information for genes assayed can be found in the Porcine Immunology and Nutritional Database (http://www.ars.usda.gov/Services/docs.htm?docid=6065).

Fluorescence signals were processed after amplification and were considered positive if the fluorescence intensity was 20-fold more or greater than the standard deviation of the baseline fluorescence. Gene expression was normalized based upon a constant amount of amplified RNA and cDNA (3, 8, 9). Relative quantification of target gene expression was evaluated by comparing linear regression lines constructed using C_T values from cDNA processed at different times for control pigs and pigs given different probiotic treatments after normalization with the RPL32 housekeeping gene. Changes in gene expression over time are determined by comparing differences in the slope of the regression lines that represented each treatment. A negative slope will indicate an upregulation of the gene since the C_T value is inversely correlated to the amount of gene expressed in the original sample.

Statistical analysis.

Bb12 tuf copy numbers in the fecal samples or intestinal contents collected from the proximal colon were determined by quantitative detection of tuf gene copies at different times posttreatment. C_T values generated after real-time PCR amplification with the Bb12 tuf gene assay were transformed to log₁₀ base to represent bacterial copy numbers of Bb12 using the tuf gene as the bacterial marker (mean ± standard error [SE]). Contrast statements were run to compare the effect of the treatment group within a time interval of treatment. Statistical results are noted for mean values from each treatment group and at each time interval evaluated, and significant differences were reported at a P value of <0.05. The analysis was performed using an SAS software package (Statview 5.0 for Macintosh Abacus Concepts, Berkeley, CA). Gene expression changes in TLR across time in proximal colon were done by a two-way analysis of covariance. Linear regression lines representing each treatment were compared to those of the control group. For these contrasts, significant differences are reported when P values are <0.05. All statistical analyses were performed using SAS Proc Mixed (SAS Institute, Inc.).

RESULTS

Species-specific real-time PCR.

A fixed amount of 100 ng of bacterial DNA extracted from pure cultures of reference strains, listed in Table 1, was used as a template for the initial validation of specificity of the duplex 5′ nuclease assay. The sequences of the designed primers and probes are listed in Table 3. Using conserved and nonconserved areas of the transaldolase gene in Bifidobacterium species, two sets of assays were designed. The transaldolase group assay identified Bifidobacterium suis, Bifidobacterium breve, and Bifidobacterium longum with C_T values less than 27 cycles, and the transaldolase assay distinguished B. animalis from all the former Bifidobacterium species with C_T values less than 31 cycles (Table 4). The GenBank-deposited sequences for the atpD, dnaK, 16S to 23S, hsp60, groES, recA, and tuf genes for Bifidobacterium-related species were also designed and tested as specific gene assays for B. animalis subspecies. The dnaK and atpD real-time PCR assays identified B. animalis subspecies but were not specific since they also reacted at a lower sensitivity with B. breve (C_T = 34) and B. suis (C_T = 36) (Table 4).

TABLE 4.

Average C_T values obtained with 100 ng of extracted DNA

Source of DNA sample	Strain	CT value for:
		Transaldolase group gene	Transaldolase gene	atpD	dnaK	16S to 23S	hsp60	groES	recA	tuf
B. animalis subsp. lactis	Ch.Hansen-Bb12	40	27	24	24	29	26	28	26	27
B. animalis subsp. lactisa	ATCC 27536	40	26	23	24	29	25	27	24	26
B. animalis	ATCC 25527	40	31	25	40	30	27	27	26	40
B. suis	ATCC 27531	27	40	36	40	40	40	40	40	40
B. breve	ATCC 15700	25	40	33	34	40	40	40	40	40
B. longum	ATCC 15707	27	40	40	40	40	40	40	40	40
E. faecium	Ch.Hansen-SF273	40	40	40	40	40	40	40	40	40
L. paracasei	Ch.Hansen-Lc-01	40	40	40	40	40	40	40	40	40
L. bulgaricus	Ch.Hansen-LBA40	40	40	40	40	40	40	40	40	40
L. acidophilus	ATCC 53544	40	40	40	30	40	40	40	40	40
Treated pig		40	32	29	30	37	31	29	31	32
Control pig		40	40	35	40	40	40	40	40	40

^aOriginally identified as B. animalis.

Regardless of the TaqMan minor groove binding probes designed for improving specificity, the 16S to 23S, hsp60, groES, and recA real-time PCR assays detect Bifidobacterium species (B. animalis and B. animalis subsp. lactis) but were not able to discriminate at the subspecies level (C_T less than 30 cycles) (Table 4). Only the tuf gene real-time PCR assay was able to differentiate B. animalis subsp. lactis (Ch.Hansen-BB12 strain; GenBank accession no. ATCC 27536) from B. animalis subsp. animalis (ATCC 25527) with C_T values less than 27 cycles. Therefore, we tested and standardized the conditions of the tuf gene real-time PCR assay (300 nM forward primer, 300 nM reverse primer, and 50 nM probe) for identification of B. animalis subsp. lactis in Bb12-treated pigs. Fecal samples from Bb12-treated pigs and non-probiotic-exposed pigs were also used as positive and negative controls for the detection of B. animalis subsp. lactis. With the exception of the transaldolase group assay, there was a positive C_T value with DNA isolated from Bb12-treated pigs (Table 4).

Sensitivity of the tuf gene assay for detection of single copy of Bb12.

The standard curve obtained from a serially diluted standard as described in Materials and Methods was linear over at least six orders of magnitude (r² = 0.99). The results indicated that the tuf gene real-time PCR assay has sensitivity to 10 copies of the gene when the tuf bacterial fragment is diluted in PBS. Smaller amounts can be detected, but the efficiency of the reaction (0.99) was substantially compromised at <10 tuf gene copies. Quantitative detection of Bb12 nucleic acid in intestinal samples spiked with lyophilized Bb12 or bacterial cultures also indicated a linear pattern (r² = 0.99) but with a lower sensitivity since the Bb12 tuf assay detected 100 copies of the tuf gene from the originally spiked material with an efficiency greater than 89%.

Quantitative detection of Bb12 in GITs of neonatal pigs.

For experiment 1, proximal colon contents were aseptically collected from all piglets on day 32 after birth. Bb12 tuf gene copies in 100 ng of DNA extracted per gram of proximal colon contents were detected at a significantly higher number (P < 0.05) in piglets from the T/T (n = 18) and C/T (n = 21) groups (1.40 ± 0.0.14 and 1.05 ± 1.3, respectively) than the T/C (n = 16) (0.36 ± 0.08) and C/C (n = 19) (0.03 ± 0.08) groups (Fig. 1). Bb12 tuf gene copies were undetectable in the intestinal contents from a separate group of control untreated piglets maintained in a different section of the farrowing barn (n = 9) (data not shown).

FIG. 1.

Bifidobacterium animalis subsp. lactis (strain Bb12) expressed as a tuf gene copy number detected in piglets. Intestinal contents were taken from the proximal colon of piglets from experiment 1 at day 32 after birth and following treatment with either Bb12 (1.05 × 10¹⁰ CFU/day) or an equivalent volume of placebo. Bacterial copy numbers were extrapolated from a standard curve for tuf gene copy number run in parallel with DNA extracted from intestinal contents as described in Materials and Methods. Data were transformed to log₁₀ base to represent bacterial copy numbers of Bb12 using the tuf gene as the bacterial marker (mean ± SE). Gray bars, control placebo-treated sow and control placebo-treated piglet (n = 19) (C/C); diagonal bars, control placebo-treated sow and Bb12-treated piglet (n = 21) (C/T); horizontal bars, Bb12-treated sow and control placebo-treated piglet (n = 16) (T/C); solid bars, Bb12-treated sow and Bb12-treated piglet (n = 18) (T/T); and empty bars, control untreated sows and untreated piglet (n = 9). Statistically significant results (P < 0.05) are noted above mean values (asterisk) of tuf gene copy number in different treatment groups.

For experiment 2, fecal samples were collected from five piglets in each of the experimental groups (C/C, C/T, T/C, and T/T) at days 10, 23, 32, and 45 after birth. All pigs were weaned on day 23, and no further Bb12 or placebo treatments were given to any of the pigs. Ten days after birth and after continuous daily treatment with either placebo or Bb12, pigs from the T/T and C/T groups had a significantly higher number of Bb12 tuf gene copies per gram of feces (2.32 ± 1.04 and 2.48 ± 1.05, respectively) compared to pigs from the T/C (0.71 ± 0.39) and C/C (0.036 ± 0.080) groups. There was, however, a significant reduction of at least 1 log in Bb12 tuf gene copies in the T/T (0.9 ± 0.6) and C/T (1.36 ± 0.70) pigs at day 23 after birth even with continuous treatment with the probiotic (P < 0.05), while Bb12-tuf gene copies in the C/C or T/C groups were lower than the linear detection limit of the assay (0.7 ± 0.89 and 0.2 ± 0.4, respectively). At day 32 after birth and 9 days after termination of treatment with Bb12, the level of Bb12 tuf gene copies was lower than the linear detection limit of the assay for all treatment groups (Fig. 2). No Bb12 tuf gene signal was detected 45 days after birth or 22 days after termination of Bb12 treatment. For experiment 3, piglets from each of the treatment groups (T/T, T/C, C/T, and C/C) were euthanized on days 7, 19, 32, or 91 after continuous treatment with either Bb12 (1.76 × 10¹⁰ CFU/pig/day) or the placebo. Bb12 tuf gene copies per gram of intestinal contents were detected in the highest numbers (a mean log range from 2.01 to 2.4/gram) in piglets from the T/T group compared to those from the C/C group throughout the entire period of examination (P < 0.05) (Fig. 3). Higher Bb12 tuf gene copies (a mean log range from 1.50 to 1.79/gram) were detected in piglets from the C/T group at days 19 and 91 than in the control (C/C) group (P < 0.05) (Fig. 3). In contrast, Bb12 tuf gene copy numbers were marginally above the level of detection in the proximal colon contents of piglets from the T/C group (a mean log range from 0.58 to 1.12/gram) or were not present in the C/C group (a mean log range from 0.08 to 0.9/gram). No difference in total bacterial load, measured by the 16S to 23S gene against all eubacteria or Lactobacillus spp., were detected in the proximal colon contents among the various treatment groups throughout the 91 days of the study (data not shown). The total number of Bifidobacterium spp., however, was significantly increased only in the T/T group (C_T value = 24.9 ± SE 1.3) compared to the C/C group (30.2 ± SE 1.74) at day 91 after continuous Bb12 treatment (P < 0.05).

FIG. 2.

Bifidobacterium animalis subsp. lactis (strain Bb12) in fecal samples from pigs following termination of oral treatment with Bb12. Bb12 expressed as a quantitative tuf gene copy number was determined in fecal samples collected at 10, 23, 32, and 45 days after birth. Piglets in the C/T and T/T groups from experiment 2 received Bb12 (1.05 × 10¹⁰ CFU/day) from birth until weaning at day 23 after birth. Bacterial copy numbers were extrapolated from a standard curve for tuf gene copy number run in parallel with DNA extracted from fecal samples as described in Materials and Methods. Gray bars, control placebo-treated sow and control placebo-treated piglet (n = 5) (C/C); diagonal bars, control placebo-treated sow and Bb12-treated piglet (n = 5) (C/T); horizontal bars, Bb12-treated sow and control placebo-treated piglet (n = 5) (T/C); and solid bars, Bb12-treated sow and Bb12-treated piglet (n = 5) (T/T). Statistically significant results (P < 0.05) are noted above mean values (asterisk) of tuf gene copy number in different treatment groups.

FIG. 3.

Bifidobacterium animalis subsp. lactis (strain Bb12) in pigs after continuous treatment with Bb12. Bb12 tuf gene copy number was determined in proximal colon contents of pigs of experiment 3 at 7, 19, 32, and 91 days after birth and continuous treatment with either 1.76 × 10¹⁰ CFU/pig/day of Bb12 or an equivalent amount of placebo. Bacterial copy numbers were extrapolated from a standard curve for tuf gene copy number run in parallel with DNA extracted from proximal colon contents as described in Materials and Methods. Gray bars, control placebo-treated sow and control placebo-treated piglet (n = 13) (C/C); diagonal bars, control placebo-treated sow and Bb12-treated piglet (n = 13) (C/T); horizontal bars, Bb12-treated sow and control placebo-treated piglet (n = 17) (T/C); and solid bars, Bb12-treated sow and Bb12-treated piglet (n = 14) (T/T). Statistically significant (P < 0.05) results are noted above mean values (asterisk) of tuf gene copy number in different treatment groups.

Local host immune response.

Mucosal tissue sections from the proximal colon were collected from all pigs in experiment 3. The levels of mRNA (C_T value) for TLR2, TLR4, and TLR9 were compared among treatment groups. Linear regression lines were constructed with the C_T values obtained at different times (days 7, 19, 32, and 91), and slopes of lines were compared among treatment groups. A negative slope indicated upregulation of the gene. Among the TLR molecules evaluated, only TLR9 was significantly upregulated over time (slope = −0.89; R2 = 0.55) in pigs from the T/T group (P < 0.01) (Table 5).

TABLE 5.

Gene expression analysis in proximal colon mucosa of pigs treated with Bb12

TLR	Slope for treatment groupa:
	C/C	C/T	T/C	T/T
2	−0.03	0.12	0.09	−0.28
4	−0.03	0.04	−0.006	−0.16
9	0.26	0.12	−0.04	−0.89b

^aSlopes of linear regression lines constructed with C_T values obtained for mRNA expression at different times (days 7, 19, 32, and 91). Changes in gene expression over time are determined by comparing differences in the slope of the regression lines that represented each treatment. A negative slope will indicate an upregulation of the gene since the C_T value is inversely correlated to the amount of gene expressed in the original sample.

^bStatistical significance is noted when P was <0.05.

DISCUSSION

To detect the presence of Bifidobacterium animalis subsp. lactis strain Bb12 in the intestinal contents of young pigs after oral ingestion, a duplex 5′ nuclease assay using the tuf gene was designed, optimized, and validated. The tuf gene is unique among the bacterial housekeeping genes tested because it is expressed as a single copy in the bacterial genome and is useful for differentiating the closely related B. animalis subsp. animalis from B. animalis subsp. lactis strain Bb12 (Table 4). With this highly specific and sensitive assay, it was possible to detect more than 100 bacterial copies of the tuf gene per 100 ng of DNA extracted from 1 g of intestinal contents from the proximal colon and feces of pigs fed viable Bb12. The number of Bb12 organisms detected using the tuf gene assay was dependent on the dietary treatment regimen and the time of sampling. Piglets treated with Bb12 after birth (C/T group) (log 1.05 ± 1.3) and those from Bb12-treated sows that were also treated from birth (T/T group) (log 1.40 ± 0.14) had the highest number of tuf gene copies in the proximal colon contents after 32 days of continuous daily feeding compared to placebo-treated pigs (C/C group) (log 0.3 ± 0.08) and placebo-treated pigs from Bb12-treated sows (T/C group) (log 0.04 ± 0.08) (Fig. 1). Similar differences were evident in fecal samples analyzed 10 days after birth (Fig. 2); Bb12 tuf gene copies detected in pigs from the C/T group (log 2.48 ± 1.05) and T/T group (log 2.32 ± 1.04) were 1 and 2 bacterial logs higher than those detected at day 23 (C/T group = 1.36 ± 0.7; T/T group = 0.85 ± 0.57). Fecal samples taken 11 (32 days after birth) and 24 (45 days after birth) days after Bb12 feeding ceased were below the detection limit, suggesting that a continuous administration of Bb12 is needed to maintain detectable levels of Bb12 in the GITs of growing pigs.

The experimental design used in these studies provided a scheme to evaluate the contribution of the sow to Bb12 numbers found in the GITs of its offspring. On days 7, 19, 32, and 91 after birth, piglets from the T/T group had significantly higher numbers of Bb12 in their proximal colons than piglets in the control group (P < 0.05) (Fig. 3). A significant difference in bacterial log numbers between piglets in the C/T group and the C/C group was only observed at day 19 after birth (1.46 ± SE 0.45) and 72 days after weaning (day 91 after birth) (1.79 ± 0.06) (P < 0.05), with a slight difference among treatment groups C/T and T/T at 91 days after treatment (P < 0.1). The Bb12 feeding regimens tested did not influence the total number of Eubacteria and Lactobacillus species (data not shown) in the proximal colon over 91 days of feeding with or without feeding Bb12 to the sow (data not shown); however, the total number of Bifidobacterium species in the T/T group at day 91 was increased. This suggests that daily feeding of greater than 10¹⁰ CFU/pig/day of Bb12 did not significantly alter other representative bacterial populations in the intestine, but Bb12 through the mother can influence the bacterial load of Bifidobacterium species in the offspring later in life. Similar observations of an influence of maternal Bifidobacterium counts on infants' fecal Bifidobacterium levels have been recently described in humans (11).

The impact of Bb12 treatment on localized host tissue responses was evaluated in the proximal colon mucosa. The TLRs are type I transmembrane proteins expressed by many cell types, including intestinal epithelial cells and classical immune cells, that function as recognition receptors for the innate immune system through binding of ligands associated largely with microbial, viral, and certain protozoan and metazoan pathogens (2, 51). Gene expression analysis on proximal colon mucosa indicated a significant upregulation of TLR9 only in pigs from the T/T group that maintained the highest number of Bb12 tuf gene copies throughout the experiment (T/T group) (P < 0.01) compared to pigs from all other groups (C/T, T/C, and C/C group) (Table 5; Fig. 3) No significant changes in mRNA expression were seen with either TLR2 or TLR4, suggesting a selective induction of TLR9 by the higher levels of Bb12. TLR9 plays a key role in the detection of bacterial DNA to induce innate immunity that is linked to the activation of various cell types for development of an acquired immune response (18, 20-22). Tohno et al. (45) reported the expression of TLR2 and TLR9 in the mesenteric lymph nodes and ileal Peyer's patches of suckling piglets and that isolated lymphocytes were induced to proliferate and release cytokines after in vitro stimulation with CpG 2006, a TLR9 ligand, zymosan, and heat-killed lactic acid-forming bacteria. Our results support these findings and add the dimension of in vivo activation of TLR9 by feeding live Bb12.

Data generated from these experiments suggest that detection of Bb12 using a single-copy Bb12 tuf gene real-time PCR assay is a useful tool to study the relationship between localized accumulation of Bb12 in the GIT and modulation of the host innate immune response. Oral feeding of live Bb12 to sows during gestation and to their piglets at the day of birth and for 91 days after birth established higher numbers of Bb12 in the intestinal contents that induced significant upregulation of mRNA expression for TLR9. These changes were not observed in pigs that did not receive Bb12 or in those whose mothers were treated only with Bb12 during the last trimester of pregnancy (T/C group). Piglets from untreated sows that received daily oral treatment with Bb12 from the day of birth through 91 days after birth also failed to express significant changes in TLR9 (C/T group). This suggests that exposure of the mother to Bb12 influenced both Bb12 load in the piglet in the presence of continuous daily feeding and significantly affected the host innate immune system development in the ambient environment. It is not clear if these differences are based on an early inoculation of the piglet as it transverses the birth canal and perianal area of the Bb12-treated sow or if the sow provides Bb12-supportive lactogenic colostrum and milk during nursing, as has recently been suggested for humans for other Bifidobacterium species (11). One could also consider epigenetic factors that could program the piglet for greater responsiveness to Bb12 during development. The pattern of upregulation of TLR9 may suggest that Bb12 provides a “priming signal” when piglets are inoculated by the sow and throughout early development. Future studies will evaluate the level of innate and acquired immunity and the degree of inflammation when these pigs are challenged with a proinflammatory stimulus.