Detection, Characterization, and In Vitro and In Vivo Expression of Genes Encoding S-Proteins in Lactobacillus gallinarum Strains Isolated from Chicken Crops

Abstract

Thirty-eight isolates of Lactobacillus gallinarum cultured from the crops of broiler chickens were screened for the presence of genes encoding S-layer proteins. All of the isolates had two S-protein genes, which were designated Lactobacillus gallinarum S-protein (lgs) genes. One gene in each isolate was either lgsA or lgsB. The Lactobacillus isolates were further characterized by pulsed-field gel electrophoresis of DNA digests, which grouped the isolates into 17 genotypes (strains). The second gene in each of eight representative strains was sequenced and shown to differ among strains (lgsC, lgsD, lgsE, lgsF, lgsG, lgsH, and lgsI). The genome of each strain thus encoded a common S-protein (encoded by either lgsA or lgsB) and a strain-specific S-protein. The extraction of cell surface proteins from cultures of the eight strains showed that each strain produced a single S-protein that was always encoded by the strain-specific lgs gene. Two of the strains were used to inoculate chickens maintained in a protected environment which were Lactobacillus-free prior to inoculation. DNAs and RNAs extracted from the digesta of the chickens were used for PCR and reverse transcription-PCR, respectively, to demonstrate the presence and transcription of lgs genes in vivo. In both cases, only the strain-specific gene was transcribed. Both of the strains adhered to the crop epithelium, consistent with published data predicting that S-proteins of lactobacilli are adhesins. The results of this study provide a basis for the investigation of gene duplication and sequence variation as mechanisms by which bacterial strains of the same species can share the same habitat.

Lactobacilli are commonly detected in gut samples collected from animal species, especially those from rodents, pigs (reviewed in reference 35), and chickens (13, 15, 20, 24, 45). In the chicken gut, lactobacilli are present in the crop (15), the ileum (20, 24), and the ceca (45). Lactobacilli dominate the relatively simple microbiota of the chicken crop (13), where at least some strains adhere to the crop epithelial surface (8, 12). Lactobacillus gallinarum, Lactobacillus crispatus, Lactobacillus johnsonii, Lactobacillus salivarius, and Lactobacillus reuteri persist in the crop throughout the life of broilers raised under commercial farming conditions (15). The first three of these species are members of the Lactobacillus acidophilus complex, which contains six closely related species in two subgroups (11, 18, 22). DNA homology group A contains L. acidophilus (A1), L. crispatus (A2), Lactobacillus amylovorus (A3), and L. gallinarum (A4), whereas DNA homology group B contains Lactobacillus gasseri (B1) and L. johnsonii (B2).

Surface-associated crystalline protein layers, termed S-layers, have been detected on cells of strains belonging to group A of the L. acidophilus complex (GAA) (4), with each strain having two different S-protein genes. S-proteins are the individual subunits that comprise S-layers (reviewed in references 30 and 32). It has been speculated that S-layers may mediate the adherence of Lactobacillus cells to the gut epithelium. For example, the S-protein CbsA, produced by L. crispatus JCM 5810 (a chicken isolate), has been shown to bind collagens and laminins (31) and to mediate adherence to chicken tissue ex vivo (2, 31). Bacterial S-layers may also protect the cell from proteases (23) or mediate selective nutrient transport (30).

The gene and protein sequences of 10 S-proteins belonging to GAA species are currently available in GenBank. A comparison of the amino acid sequences of these proteins has revealed two conserved regions, namely, an N-terminal signal sequence of ca. 30 amino acids which directs secretion of the protein by the general secretory pathway (5) and a C-terminal domain of ca. 123 amino acids that anchors the protein to the cell surface (34). A variable domain is located between the two conserved regions. This variable domain has been shown to mediate ex vivo tissue binding, S-protein folding, and crystal formation (2, 31, 33, 34).

For this study, we sequenced the variable region of eight different S-protein genes of L. gallinarum isolates. The sequence information was used to determine the prevalence, using PCR and DNA-DNA hybridization, of specific S-protein genes in a collection of L. gallinarum strains. We then determined which S-protein was produced by representative strains when cultivated in laboratory medium or when inhabiting the guts of chickens maintained in a protected environment. Our results provide a basis for the study of S-protein genes in relation to the microbial ecology of the crop. Future studies may provide an understanding of how closely related Lactobacillus species, as well as strains of the same species, can coexist in the crops of chickens.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The bacterial strains and plasmids used in this study are described in Table 1. The L. gallinarum isolates were cultured from chicken crops as described by Guan et al. (15) and were stored at −80°C in Lactobacilli MRS Broth (Difco, Becton Dickinson, Sparks, MD) containing 50% glycerol. The identification of Lactobacillus species was done as described previously (15). Lactobacilli were propagated using Lactobacilli MRS Agar at 37°C for 48 h under anaerobic conditions (5% CO₂, 10% H₂, and the balance N₂). Escherichia coli JM109 (Table 1) (27) was grown at 37°C with aeration in 2YT broth or on LB agar plates (29). To facilitate the selection of transformants, ampicillin (100 μg/ml) was used to supplement LB and 2YT media, and 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal, 80 μg/ml) and isopropyl-β-d-thiogalactopyranoside (IPTG, 0.5 mM) were added to LB agar plates.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristics	Reference or source
Bacterial strains
L. gallinarum
D41, D109,a D148	Crop isolate, PFGE group 1	This study
D45, D71, D73, D75, D79, D80, D108a	Crop isolate, PGFE group 2	This study
D149, D195,a D197	Crop isolate, PFGE group 3	This study
D42,a D47	Crop isolate, PFGE group 4a	This study
D46	Crop isolate, PFGE group 4b	This study
D44a	Crop isolate, PFGE group 5	This study
D260a	Crop isolate, PFGE group 6	This study
15-5, 15-8	Crop isolate, PFGE group 7	This study
28-1, 28-6, 28-7	Crop isolate, PFGE group 8	This study
28-3, 28-4, 28-9, 31-2, 31-5, 31-7, 31-9	Crop isolate, PFGE group 9	This study
D255a	Crop isolate, PFGE group 10	This study
D256a	Crop isolate, PFGE group 11	This study
15-4	Crop isolate, PFGE group 12	This study
15-6	Crop isolate, PFGE group 13	This study
15-9	Crop isolate, PFGE group 14	This study
26-3	Crop isolate, PFGE group 15	This study
31-4	Crop isolate, PFGE group 16	This study
31-6	Crop isolate, PFGE group 17	This study
ATCC 33199Ta	Crop isolate, type strain	ATCCb
E. coli JM109	recA1 endA1 gyrA96 thi hsdR17 (rk−mk+), relA1 supE44 Δ(lac-proAB) [F′ traD36 proAB lacIqZΔM15]	Promega (27)
Plasmids
pGEM-T Easy		Promega
pD42C	pGEM-T::lgsH (1.2 kb)c	This study
pD44B	pGEM-T::lgsG (1.2 kb)c	This study
pD109D	pGEM-T::lgsA (1.2 kb)c	This study
pD109E	pGEM-T::lgsF (1.3 kb)c	This study
pD195F	pGEM-T::lgsB (1.2 kb)c	This study
pD195C	pGEM-T::lgsC (1.6 kb)c	This study
pD255	pGEM-T::lgsD (1.2 kb)c	This study
pD256	pGEM-T::lgsE (1.2 kb)c	This study

^aRepresentative of PFGE group from which S-protein genes were cloned and characterized.

^bATCC, American Type Culture Collection.

^cThe size of the fragment cloned into pGEM-T Easy is indicated.

Genetic fingerprinting of L. gallinarum isolates by PFGE of DNA digests.

Pulsed-field gel electrophoresis (PFGE) plugs were prepared and digested (SmaI) following the protocol of Tanskanen et al. (38), as modified by Walker and Klaenhammer (42). This procedure was further modified as follows: 1.5 ml or 9 ml of log-phase culture at an optical density at 600 nm of 0.6 or 0.3, respectively, was used to prepare the agarose plugs, which were treated with proteinase K prior to the mutanolysin/lysozyme step to ensure cell lysis. Samples and molecular weight markers were run in 1% agarose gels in 0.5× TBE buffer (0.045 M Tris-borate, 0.001 M EDTA, pH 8.3) at 14°C using a Bio-Rad CHEF Mapper XA (Hercules, CA) under the following conditions: 6.0 V/cm, 2.98 s to 17.33 s switch time, linear ramping factor, 120° angle, and a run time of 26 h 56 min. Gels were analyzed visually, using the strain differentiation criteria described by Tenover et al. (39), and in silico, using Bionumerics software (Applied Maths, Austin, TX) to derive Dice's similarity coefficient values with a Ward dendrogram using fuzzy logic and 6% position tolerance.

Cloning and sequencing of S-protein genes.

DNA was extracted from L. gallinarum isolates using the protocol of Walter et al. (44). Total DNA was diluted 1/20 (vol/vol), and 1 μl (∼100 ng/μl) was used as template DNA in an S-protein gene PCR mix that also contained a 0.2 mM concentration of each deoxynucleoside triphosphate (Invitrogen, Burlington, Ontario, Canada), 1 mM MgCl₂, 1 pmol/μl of each oligonucleotide (Invitrogen), and 0.05 U/μl Taq polymerase (Invitrogen) in a total volume of 20 μl. An alignment of the previously sequenced GAA S-protein genes cbsA, cbsB, slpnA, slpnB, slpA, and slpB (NCBI accession numbers AF001313 [31], AF079365 [31], AF253043 [31], AF253044 [31], X89375 [5], and X89376 [5], respectively) was used to design degenerate primers that were used to screen for S-protein genes in L. gallinarum. Primers Usl-1 (5′-GAATYGTKAGCGCTSCTGCTGC-3′, corresponding to nucleotides [nt] 17 to 38 of cbsA) and Usl-2 (5′-GTAAACGTAWGCGTTGTGCTTC-3′, corresponding to the complement of nt 1170 to 1149 of cbsA) were designed to amplify a DNA fragment that encoded part of the conserved N-terminal signal sequence (ca. 18 amino acids), the entire N-terminal variable region (ca. 280 amino acids), and approximately half of the C-terminal conserved domain (ca. 60 amino acids). Amplification of the L. gallinarum S-protein genes utilized a touchdown PCR program in a GeneAmp PCR System 9700 thermocycler (Applied Biosystems, Foster City, CA) consisting of the following: 5 min at 94°C; 2 cycles of 94°C for 30 s, 63°C for 30 s, and 72°C for 1 min; 2 cycles of 94°C for 30 s, 62°C for 30 s, and 72°C for 1 min; 2 cycles of 94°C for 30 s, 61°C for 30 s, and 72°C for 1 min; 30 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 1 min; and 7 min at 72°C. PCR products were analyzed in a 1% agarose gel. Selected PCR products were purified using a Qiaex II kit (QIAGEN, Mississauga, Ontario, Canada), ligated into the pGEM-T Easy vector, and transformed into competent E. coli JM109 cells following the manufacturer's instructions (Promega, Madison, WI). After transformation, E. coli cells were grown in SOC broth (29) for 1.5 h and plated on LB agar containing ampicillin, X-Gal, and IPTG (29).

Plasmid DNAs were extracted from E. coli transformants using an alkaline lysis method (29), purified using polyethylene glycol precipitation (29), and initially sequenced with T7 and Sp6 oligonucleotides using an ET dye terminator kit (Amersham Biosciences, Baie D'Urfe, Quebec, Canada). Primer walking was performed to complete the sequences of the cloned inserts. All of the sequencing reactions were carried out by the Agricultural Genomics and Proteomics Unit or the Biological Sciences Molecular Biology Service Unit at the University of Alberta. Sequences were assembled and aligned with the Lasergene software package (DNAStar Inc., Madison, WI) and/or Vector NTI Advance 9.0 (Informax, Frederick, MD). Consensus sequences were determined by comparing sequences from both DNA strands.

S-protein extraction and characterization.

Extracellular proteins were extracted from Lactobacillus cells using a protocol modified from the work of Chagnaud et al. (9). Briefly, after overnight growth in broth, the cells were washed twice in a 0.85% sodium chloride solution and then treated with extraction buffer (2% sodium dodecyl sulfate [SDS], 1% β-mercaptoethanol) at 70°C for 10 min. The cells were pelleted by centrifugation, and the supernatant (the cell surface-associated protein extract) was analyzed by 15% SDS-polyacrylamide gel electrophoresis (SDS-PAGE; 15% acrylamide [37.5 acrylamide:1 bisacrylamide {vol/vol}], 0.1% SDS [wt/vol], 0.375 M Tris-HCl, pH 8.8) in a Bio-Rad Mini-PROTEAN 3 system (Bio-Rad, Hercules, CA). The gel was run at 100 V for 3.5 h and stained with Sigma EZBlue staining reagent (Sigma, Oakville, Ontario, Canada) according to the manufacturer's directions.

To identify which S-proteins were produced in vitro by the L. gallinarum strains, S-protein bands from representative isolates were excised from the polyacrylamide gel and subjected to quadrupole-time of flight (Q-TOF) analysis by the Institute for Biomolecular Design at the University of Alberta.

Phylogenetic protein analysis was performed by parsimony using PAUP 4.0 (Sinauer Associates, MA).

Screening of L. gallinarum isolates for specific S-protein genes.

PCRs using the S-protein gene-specific oligonucleotides listed in Table 2 were used to screen L. gallinarum isolates. An alignment of the gene sequences was used to identify unique regions to which gene-specific oligonucleotides were targeted for lgsA, lgsB, lgsC, lgsE, lgsF, lgsG, lgsH, and lgsI, but attempts to design primers specific for lgsD were unsuccessful. A reaction containing 1 μl of diluted template DNA (prepared as outlined above), 0.2 mM deoxynucleoside triphosphate mix, 1 mM MgCl₂, 1 pmol/μl each oligonucleotide, and 0.05 U/μl Taq polymerase was prepared for each isolate and amplified under the following conditions: 94°C for 5 min followed by 30 cycles of 30 s at 94°C, 30 s at the oligonucleotide-specific annealing temperature (T_a) (Table 2), and 1 min of extension at 72°C, and then a hold at 72°C for 7 min. One positive and two negative controls were included in every PCR run. The positive control consisted of 1 μl of diluted DNA of the isolate from which the gene originated (e.g., D109 for Fsl-1/Rsl-1; Table 2). The negative controls consisted of 1 μl sterile MilliQ H₂O and 1 μl diluted (1/20) L. johnsonii ATCC 33200 DNA (S-protein-negative strain). When necessary, dimethyl sulfoxide (7% [vol/vol]) was added to the PCR mixtures to ensure specificity. The presence or absence of PCR products and the sizes of the fragments for positive PCRs were analyzed using a 1% agarose gel.

TABLE 2.

S-protein gene-specific oligonucleotides used to investigate the prevalence of different genes among L. gallinarum isolates

Gene	Primer	Sequence (5′ to 3′)	Ta (°C) used for screening PCR	Size of amplified fragment (kb)
lgsA	Fsl-1	CATCAAACAATACTGTTACAAAC	51	0.34a
	Rsl-1	CAAACCATTAGCATCAATATC
lgsB	Fsl-2	GATAACGGATCACACTTTGG	53	0.33a
	Rsl-2	AGGTTGTTAATGTTAGCTGTAG
lgsC	Fsl-3	CAAAGGCTGAAATGACTACTAC	57	0.6a
	Rsl-3	CTTAGCATTGTCAACATGTACG
lgsE	Fsl-5	CAACTTGACAAGAATGGTACTG	57	0.26
	Rsl-5	GTACTTGATGTCACTTGATTCC
lgsF	D109E-2	CATTAACTTGGGTGGTTCAGACG	63	0.22a
	D109E-3	TACCGCCTTGTACAGCAGTAACC
lgsG	Fsl-6	GTTACTATCACTGTTCCAGCAGG	63	0.30
	Rsl-6	CTACGTAGTTCTTATCAAATTCAG
lgsH	Fsl-7	GGGCTGCACATAATGCGACTGA	63	0.28
	Rsl-7	GAAATAGCACTTGCATCTGAGCC
	D42C-3	ATACTAAGGCTGACTCAGTAG	64	0.58
	Rsl-7	GAAATAGCACTTGCATCTGAGCC
lgsI	P5E-1	TAGCTGATGCTGCACAAGACG	59	0.52
	P5E-3	ACTAGCATCAACATCAACACCG

^aThe expected sizes of the amplified products for the primers used in the clean-room experiments are indicated.

DNA-DNA hybridization utilizing a universal S-protein gene probe was also used to verify the number of S-protein genes in each strain. DNAs in agarose plugs prepared for PFGE were digested with EcoRI, and samples and molecular weight markers were run in 1% agarose gels at 14°C in a CHEF Mapper XA instrument under the following field inversion gel electrophoresis conditions: initial switch time of 0.06 s, final switch time of 0.08 s, forward voltage gradient of 9 V/cm, reverse voltage gradient of 6 V/cm, and a running time of 17 h 29 min. DNAs were then transferred and fixed to a nitrocellulose membrane using standard methods (29). The 186-bp universal probe, consisting of the conserved 3′ sequence of S-protein genes, was obtained by PCR amplification using DNA extracted from strain D109, the primers Usl-3 (CATACTTCTACGACAAGGACGC, corresponding to nt 956 to 986 of cbsA) and Usl-2, and the touchdown program described above. The probe was labeled using the Invitrogen Random Primer DNA Labeling System according to the manufacturer's instructions. Standard buffers (29) were used for hybridization, which was conducted at 50°C. After high- and low-stringency washes (29), the membrane was exposed to a phosphorimaging plate for 2 h to detect hybridization signals.

S-protein gene expression in the chicken gut.

Chickens were hatched and raised under “clean-room” conditions to exclude colonization of the birds by lactobacilli. The experimental protocol was approved by the University of Alberta Health Sciences Animal Policy and Welfare Committee (protocol number 329/01/05). The egg incubator (Lyon model RX2; Lyon Electric Company, Chula Vista, CA) was chemically disinfected with Virkon (Dispar, Joliette, Canada; diluted 1:100 [wt/vol] in water) solution. Ross 308 broiler eggs (Aviagen Inc., Huntsville, AL) from a commercial hatchery (Lilydale Hatchery, Edmonton, Alberta, Canada) were thoroughly sprayed with Virkon inside the chemically disinfected incubator, and the lid was sealed. Eggs were incubated between 36.9 and 38.3°C at 65% humidity for 21 days. On the day of hatching, birds were transferred to a cage that had been chemically disinfected with Clidox solution (1:5:1; base:water:activator), where they were given sterile water and gamma-irradiated feed (PMI Nutrition International) ad libitum. The temperature in the cage was maintained between 35 and 40°C, and the birds were supplied with fresh sterile bedding (shavings) daily. The incubator and cage were maintained in a HEPA-filtered environment at all times (Forma animal isolator) in a room that was maintained under positive pressure. Before and after each experiment, the whole room was cleaned with Clidox solution (1:18:1), and all equipment and supplies were cleaned with Clidox or Virkon as outlined above and left to sit for 1 week.

Lactobacillus strains were streaked on Lactobacilli MRS Agar and incubated anaerobically at 37°C for 48 h. A single colony was used to inoculate Lactobacilli MRS Broth and incubated overnight under the same conditions. The culture was pelleted (5,000 × g), washed, and resuspended in the same volume of phosphate-buffered saline (PBS, pH 7.4). The CFU/ml was determined by plating serial dilutions of the suspension on Lactobacilli MRS Agar plates and counting the CFU after 48 h of incubation. A PBS suspension (0.5 ml) containing L. gallinarum strain D109 or D195 (10⁹ cells/ml) was administered to each bird by gavage 2 days after hatching. Samples of the crop, ileum, and ceca were collected from 5 and 10 birds on days 7 and 14, respectively. Approximately 100 mg of digesta from each section of the gastrointestinal tract was collected from five birds at each sample time within 5 min of death, placed in 1 ml of RNAlater RNA stabilization reagent (QIAGEN, Hilden, Germany), and stored at room temperature during transit to the laboratory. For each experiment, samples from another five birds (day 14) were examined by bacteriological culture as described previously (15). The identity of selected colonies was confirmed by 16S rRNA gene sequencing with SacI-POMod and HDA2 primers (15). Additionally, lactic acid bacterium profiles were generated by Lac PCR-denaturing gradient gel electrophoresis (DGGE) as described previously (15), using DNAs extracted from gut samples and pure cultures.

Bacterial RNAs and DNAs were extracted from gut samples by using a QIAGEN RNA/DNA mini kit according to the manufacturer's protocol. The amount and quality of RNA and DNA were estimated by measuring the absorbance at 260 and 280 nm in a GeneQuant spectrophotometer (Fisher Scientific). RNAs were stored at −80°C, and DNAs were stored at −20°C.

The presence and expression of S-protein genes in gut samples were investigated using PCR and reverse transcription-PCR (RT-PCR) amplifications. Primers Fsl-1/Rsl-1 and D109E-2/D109E-3 (Table 2) were used to detect lgsA and lgsF, respectively, in samples obtained from birds who received strain D109; primers Fsl-2/Rsl-2 and Fsl-3/Rsl-3 (Table 2) were used to detect lgsB and lgsC, respectively, in samples from birds who received strain D195. The absence of DNA contamination in RNA samples was verified by PCR, using RNA as a template with primers Usl-1 and Usl-2, and was checked in a 2% agarose gel. RT-PCR was performed using a QIAGEN One-Step RT-PCR kit (QIAGEN) according to the manufacturer's instructions. All PCRs and RT-PCRs were conducted using the programs outlined above and the annealing temperatures indicated in Table 2. The identities of the RT-PCR products were confirmed by sequencing amplified DNA fragments.

Crop samples were prepared for scanning electron microscopy by placing a 1-cm² piece of tissue in 2% glutaraldehyde solution immediately after the crop tissue was obtained from the bird. Standard methods were used to further process samples in the Microscopy Unit of the Department of Biological Sciences, University of Alberta. Briefly, the fixed specimens were washed in PBS, dehydrated through an ethanol series, treated with hexamethyldisiazane, air dried, and sputter-coated with gold. Cryosections or paraffin-blocked sections were also prepared and stained using a tissue Gram stain and standard methods (17).

The same procedures were used in a control trial in which only sterile PBS (0.5 ml) was administered to the birds on day 2. Crop, ileal, and cecal samples were collected from each bird on day 14 and analyzed using the procedures outlined above for bacterial culture, profiling of lactic acid bacterial populations, and electron microscopy. Bacteria were not detected on the MRS plates, and fragments comigrating with lactobacilli were not observed by PCR-DGGE, thus confirming the Lactobacillus-free status of the birds.

Nucleotide sequence accession numbers.

The sequences determined in this study were deposited in the GenBank database (accession numbers AY597259 to AY597268).

RESULTS

Genetic fingerprinting of L. gallinarum isolates using PFGE.

SmaI PFGE was used to genetically fingerprint 38 L. gallinarum isolates (Table 1). Bionumerics software and visual analysis were used to compare the SmaI profiles of the isolates and revealed that there were 17 unique profiles or strains (PFGE groups 1 to 17). Representative profiles and relationships between PFGE groups are shown in Fig. 1. Slight variations in PFGE patterns were noted for PFGE group 4 and were designated by subgroupings “a” and “b.”

FIG. 1.

Relationships (Dice's similarity coefficient, or DICE correlation) between SmaI PFGE profiles of representative isolates from each strain (PFGE group). MWM, corresponding fragment sizes of the lambda ladder (New England Biolabs, Pickering, Ontario, Canada).

It was possible to relate each isolate to its bird of origin. This showed that several strains of L. gallinarum, differentiated by PFGE with DNA digests, were detected in each bird, indicating that multiple strains were able to coexist in the same crop (see Table 3 for examples).

TABLE 3.

Examples of coexistence of L. gallinarum strains in the crops of broilers

Bird no.	PFGE group (isolate no.) obtained from crop	S-protein genes
1	1 (D41)	lgsA, lgsF
	2 (D45)	lgsA, lgsH
	4a (D42)	lgsA, lgsH
	5 (D44)	lgsA, lgsG
2	1 (D109)	lgsA, lgsF
	2 (D108)	lgsA, lgsH
3	6 (D260)	lgsA, unknown
	10 (D255)	lgsA, lgsD
	11 (D256)	lgsA, lgsE
4	1 (D148)	lgsA, lgsF
	3 (D149)	lgsB, lgsC

Detection and characterization of L. gallinarum S-proteins.

Thirty-eight isolates of L. gallinarum were screened for S-layer protein-encoding genes by PCR using the universal primers Usl-1 and Usl-2. All of the isolates produced a PCR product of the expected size (∼1.2 kb). The PCR products amplified from eight L. gallinarum strains (each from a different PFGE group, as indicated in Table 1) and the type strain were cloned. Sequencing analysis of the cloned DNAs revealed eight genes that were designated lgs (L. gallinarum S-protein) followed by consecutive letters of the alphabet. The encoded proteins and their strains of origin are listed in Table 4. Two different S-protein genes were cloned from L. gallinarum D109 (lgsA and lgsF), L. gallinarum D195 (lgsB and lgsC), and ATCC 33199^T (lgsB and lgsI).

TABLE 4.

Similarities between L. gallinarum S-proteins and other Lactobacillus S-proteins

Predicted protein (Genbank accession no.)	Origin of sequence (PFGE group)	Highest identity in BLAST-P (GenBank accession no. and reference, where available)	Species	% Sequence identity
LgsA (AY597260)	D109 (1)	SlpnB (AAF68972) (31)	L. crispatus	67
		LbsB (BAC76687)	L. crispatus	65
		CbsA (AAB58734) (31)	L. crispatus	61
LgsB (AY597261)	D195 (3)	Slp (AAS83409)	L. gallinarum	99
		CbsA (AAB58734)	L. crispatus	63
		LbsB (BAC76687)	L. crispatus	59
LgsB (AY597262)	ATCC 33199T	Slp (AAS83409)	L. gallinarum	99
		CbsA (AAB58734)	L. crispatus	64
		LbsB (BAC76687)	L. crispatus	58
LgsC (AY597263)	D195 (3)	Conserved C-terminal anchor domain of Slp (AAS83409)	L. crispatus	88
		Short region of S-layer (BAB72065)a	L. helveticus	36
		Short region of S-layer (BAB72065)a	L. helveticus	26
LgsD (AY597264)	D255 (10)	LbsB (BAC76687)	L. crispatus	63
		SlpnB (AAF68972)	L. crispatus	62
		CbsA (AAB58734)	L. crispatus	60
LgsE (AY597265)	D256 (11)	SlpnB (AAF68972)	L. crispatus	84
		LbsB (BAC76687)	L. crispatus	73
		CbsA (AAB58734)	L. crispatus	68
LgsF (AY597266)	D109 (1)	CbsB (AAC28100) (31)	L. crispatus	48
		S-layer (BAB72066)	L. helveticus	44
		S-layer (BAB72065)	L. helveticus	43
LgsG (AY597267)	D44 (5)	SlpB (CAA61561) (5)	L. acidophilus	43
		Slp (AAS83409)	L. gallinarum	48
		S-layer (CAB46987) (40)	L. helveticus	44
LgsH (AY597268)	D42 (4a)	LbsB (BAC76687)	L. crispatus	59
		CbsA (AAB58734)	L. crispatus	62
		SlpnB (AAF68972)	L. crispatus	59
LgsI (AY597259)	ATCC 33199T	LbsA (BAC76686)	L. crispatus	69
		S-layer (BAB72065)	L. helveticus	59
		S-layer (BAB72066)	L. helveticus	60

^aExcluding conserved C-terminal anchor domain.

A BLAST-P (1) analysis of predicted S-protein sequences indicated that L. gallinarum S-proteins LgsA, LgsB, LgsD, LgsE, and LgsH were similar to SlpnB, LbsB, and CbsA of L. crispatus (Table 4) and related S-proteins of Lactobacillus helveticus and L. acidophilus (data not shown). LgsC was dissimilar from the other sequences but had low BLAST-P homology to the variable region of an S-protein from L. helveticus GCL1001 (Table 4). LgsB obtained from ATCC 33199^T was 99% identical to L. gallinarum Slp (GenBank accession no. AAS83409) and 97.8% identical to LgsB from D195, with eight single amino acid substitutions and six additional amino acids in the variable region. The BLAST-P results therefore confirmed that the cloned genes encoded unique S-proteins (based on <84% identity to other known S-proteins) and indicated strain-specific differences in the amino acid sequences. Parsimony analysis was used to compare the L. gallinarum S-proteins and construct an unrooted dendrogram (Fig. 2). LgsA, LgsB, LgsD, LgsE, and LgsH formed a cluster, but the remaining Lgs proteins were distantly related.

FIG. 2.

Unrooted dendrogram constructed by parsimony analysis of the amino acid sequences of Lgs proteins. Bootstrap values for 1,000 trials are shown at major internal nodes.

Prevalence of S-protein genes among L. gallinarum isolates.

Based on PCR detection of S-protein genes, 34 of the isolates harbored lgsA, 4 harbored lgsB, 3 harbored lgsC, 1 harbored lgsE, 7 harbored lgsF, 1 harbored lgsG, 10 harbored lgsH, and 1 harbored lgsI (Table 5). The PCR screening results were confirmed by cloning and sequencing lgsA from D260, D255, and D256 and lgsB from D149, D195, and D197. lgsA and lgsB were common among the L. gallinarum isolates but were mutually exclusive (Table 5). In strains for which a second gene (not lgsA or lgsB) was detected, all members of the same strain (PFGE profile) had the same gene, but between strains the second S-protein gene was usually different (Table 5). PCR screening did not detect the second gene in 14 isolates representing seven PFGE groups (Table 5). A representative isolate from each of these groups was tested in a DNA-DNA hybridization experiment using a universal S-protein gene probe. Two DNA fragments from each strain hybridized to the probe, indicating that there were two S-protein genes per strain (data not shown).

TABLE 5.

Prevalence of S-protein genes among L. gallinarum strains and their expression by representative isolates in vitro

PFGE group (no. of isolates screened)a	Presence or absence of S-protein gene									S-protein expressed in vitro (isolate analyzed)
	lgsA	lgsB	lgsC	lgsDb	lgsE	lgsF	lgsG	lgsH	lgsI
1 (3)	+	−	−	NT	−	+	−	−	−	LgsF (D109)
2 (7)	+	−	−	NT	−	−	−	+	−
3 (3)	−	+	+	NT	−	−	−	−	−	LgsC (D195)
4a (2)	+	−	−	NT	−	−	−	+	−	LgsH (D42)
4b (1)	+	−	−	NT	−	−	−	+	−
5 (1)	+	−	−	NT	−	−	+	−	−	LgsG (D44)
6 (1)	+	−	−	NT	−	−	−	−	−
7 (2)	+	−	−	NT	−	−	−	−	−
8 (3)	+	−	−	NT	−	+	−	−	−
9 (7)	+	−	−	NT	−	−	−	−	−
10 (1)	+	−	−	+	−	−	−	−	−	LgsD (D255)
11 (1)	+	−	−	NT	+	−	−	−	−	LgsE (D256)
12 (1)	+	−	−	NT	−	−	−	−	−
13 (1)	+	−	−	NT	−	+	−	−	−	LgsF (15-6)
14 (1)	+	−	−	NT	−	−	−	−	−
15 (1)	+	−	−	NT	−	−	−	−	−
16 (1)	+	−	−	NT	−	−	−	−	−
17 (1)	−	+	−	NT	−	−	−	−	+
ATCC 33199T	−	+	−	NT	−	−	−	−	+	LgsI

^aIsolates belonging to each PFGE group are described in Table 1.

^bNT, not tested.

In vitro expression of L. gallinarum S-proteins.

The extracellular protein profiles showed that each strain produced a dominant protein ranging in size from 43 to 52 kDa (Fig. 3), which was consistent with the sizes of S-proteins produced by other GAA species. Isolates belonging to the same strain (PFGE group) produced S-proteins of the same size (data not shown). To determine which protein was produced in vitro, the S-protein bands were excised from the SDS-PAGE gels for the eight strains and subjected to Q-TOF analysis. The protein sequencing results agreed with the protein predicted from the gene sequence of each strain. Neither LgsA nor LgsB was produced in vitro, but all strains expressed the alternative, strain-specific gene (Table 5).

FIG. 3.

Extracellular protein profiles of representative L. gallinarum isolates from each PFGE group analyzed by 15% or 10% (for ATCC 33199^T only) SDS-PAGE. Isolate numbers and their corresponding PFGE groups are indicated. MWM, broad-range molecular size marker (Bio-Rad, Hercules, CA), with corresponding sizes of the protein bands indicated.

In vivo expression of L. gallinarum S-proteins.

Given that all 38 isolates contained either lgsA or lgsB, both of which are conserved yet mutually exclusive genes (Table 5), we hypothesized that the LgsA and LgsB S-proteins may confer an important quality among the strains, such as persistence in the gastrointestinal tract. Therefore, we chose strains D109 (lgsA lgsF) and D195 (lgsB lgsC) for in vivo experiments. Quantitative culturing of lactobacilli from the digesta of inoculated birds showed that both strains achieved population levels of about 10⁷ CFU per gram in the crop, ileum, and ceca. The 16S rRNA genes from subcultures of small opaque colonies recovered from the birds were sequenced and had 100% nucleotide sequence identity to D109, D195, and the L. gallinarum type strain (GenBank accession no. AJ417737). In both experiments, large white colonies were present in about equal numbers to Lactobacillus colonies in cultures prepared from ileal and cecal samples but were approximately 100-fold fewer in the crop (data not shown). The sequences of the 16S rRNA genes of these colonies were found to have 100% identity to Pediococcus acidilactici (GenBank accession no. AF515229 [10]). When DNAs obtained from these colonies were used as templates for PCRs with primers specific for lgsA, lgsF, lgsB, and lgsC, amplification products were not detected (data not shown). The culture-based observations were further confirmed using Lac PCR-DGGE, which showed that strains D109 and D195 were present in the crop, ileum, and ceca of each bird (Fig. 4). A band migrating to the same position as P. acidilactici was only present in the profiles generated from ileal and cecal samples (Fig. 4). Scanning electron microscopy and light microscopy of crop specimens revealed the presence of rod-shaped bacteria adherent to crop epithelial cells (Fig. 5B and C). These bacteria were not seen in preparations obtained from control birds inoculated in a separate experiment with PBS alone (Fig. 5A).

FIG. 4.

PCR-DGGE profiles (30 to 45% denaturing gradient gel) generated from DNAs extracted from gut samples of chickens kept under clean-room conditions (clean-room birds) using primer pair Lac1 and Lac2-GC. (A) Profiles for chickens who received L. gallinarum strain D109 by gavage. (B) Profiles for chickens who received L. gallinarum strain D195 by gavage. The bird (designated by C29, C30, etc.) and gut region (Cr, crop; Il, ileum; Ce, cecum) of origin are indicated at the top of each gel. D109 and D195, PCR products from pure cultures of L. gallinarum strains; SO, PCR products from small opaque Lactobacillus colonies isolated from clean-room birds; LW1 and LW2, PCR products from large white P. acidilactici colonies isolated from birds in D109 and D195 clean-room experiments, respectively.

FIG. 5.

Microscopic analysis of samples from clean-room birds. Scanning electron micrographs show the crop epithelial surface of a PBS-inoculated bird (A) and bacterial cells adherent to crop epithelium from a bird inoculated with L. gallinarum D109 (B). (C) Light micrograph of bacterial cells associated with the crop epithelium of a bird inoculated with L. gallinarum D195.

PCR-amplified fragments of the expected size for lgsA and lgsF (Table 2) were obtained from template DNAs extracted from birds who received strain D109 by gavage (Fig. 6A and B; data not shown). When RNAs extracted from these birds were used as templates for RT-PCR, however, amplified fragments were only obtained from reactions containing the lgsF primers (Fig. 6C to F). Similarly, both lgsB and lgsC were amplified in PCRs containing DNAs extracted from D195-inoculated birds, but only lgsC was amplified by RT-PCR using RNAs obtained from the crop, ileum, and ceca (data not shown). The identity of lgsF and lgsC was confirmed by sequencing the RT-PCR products. Therefore, for both strains, only the strain-specific lgs genes, lgsF and lgsC, were expressed in the guts of chickens as well as in laboratory culture. Neither lgsA nor lgsB was expressed in vitro or in vivo.

FIG. 6.

Presence and expression of S-protein genes in the guts of clean-room birds inoculated with L. gallinarum strain D109. (A) PCR detection of lgsA in crops of birds A to J. (B) PCR detection of lgsF in crops of birds A to J. (C) RT-PCR detection of expression of lgsF in the crops of birds A to D, with a lack of expression of lgsA in the crops of birds A to D. (D) RT-PCR indicates a lack of expression of lgsA in the crops of birds F to J but the detection of expression of lgsF in the crops of birds F to J. (E) RT-PCR indicates a lack of detection of lgsA expression in the ilea and ceca of birds A to J. (F) RT-PCR detection of expression of lgsF in ilea and ceca of birds A to J. Birds A to D were 7 days old; birds F to J were 14 days old. Cr, crop; Il, ileum; Ce, cecum; 33199, L. gallinarum type strain ATCC 33199 (negative control); 33200, L. johnsonii type strain ATCC 33200 (negative control); Ne, negative control containing water only; M, molecular weight marker.

DISCUSSION

Lactobacilli can be consistently detected in the guts of chickens raised under commercial conditions (15, 20, 24, 45). These lactobacilli are “autochthonous” to the guts of the birds because they become established soon after the birds hatch, form biofilms on the crop epithelium, have demonstrable in vivo activity, and persist throughout the life of the host at predictable population levels in particular regions of the gut (36). We searched for and characterized genes encoding S-proteins in L. gallinarum isolates cultured from the crops of broilers. In vivo observations have not yet been reported that define the role of these S-proteins in the microbial ecology of the crop, although they have been reported to mediate adherence to chicken cells ex vivo (2, 31).

We observed that, as reported for other GAA lactobacilli, two S-protein genes were present in each strain of L. gallinarum. Two common but mutually exclusive S-protein genes were detected among the L. gallinarum isolates, i.e., lgsA and lgsB. Moreover, the expression of these genes was not detected during growth of the bacteria in laboratory media or in the guts of chickens. In contrast, each strain harbored a second S-protein gene that was mostly strain specific (lgsI, lgsF, and lgsH were exceptions) and which was expressed both in vitro and, for the two strains that we tested, in vivo. These results imply that, in general, the diversity of strains of L. gallinarum that coexist in the crops of broilers each produces S-proteins with differing amino acid compositions.

The surface diversity detected in bacteria that are gut residents has been described as a mechanism to evade the host immune response that might be directed at the bacterial cell surface antigens. For example, Bacteroides fragilis NCTC 9343 modulates its cell surface antigenicity by producing at least eight distinctly different capsular polysaccharides. This is achieved by regulating the expression of polysaccharide-encoding loci in an on-off manner by the reversible inversion of DNA segments that contain the promoters for their expression (21). S-layer variation due to the expression of alternative S-protein genes is well known and is usually due to DNA rearrangements (reviewed in reference 7). For example, L. acidophilus ATCC 4356 harbors two S-protein genes, one of which (slpA) has favored expression in laboratory culture (99.7% of cells). The second S-protein gene (slpB) is located 6 kb downstream, but in the opposite orientation. In a small proportion of cells in the culture, the segment is inverted, placing the silent gene behind the S promoter, which is located upstream of the inverted segment (3-6). Environmental conditions could select for cell populations expressing the S-protein most appropriate for the habitat. Hence, we reasoned that a similar method of gene regulation, if it existed in L. gallinarum, could result in the expression of alternative S-protein genes in the guts of chickens compared to those expressed in culture medium. The gut environment, however, still favored the expression of the strain-specific genes that were expressed in laboratory culture, at least in the case of the two strains that we tested.

The cecal ecosystem of chickens, as is the case for other animal species, is inhabited by a huge diversity of bacterial phylotypes (45). The diversity associated with gut bacterial communities can be explained in terms of the “niche exclusion principle,” which states that a single ecological niche can support no more than one type of organism, either genotype or species (16). Hence, in the cecum, where heterogeneous environments exist in relation to bacterial nutrients and spatial arrangements, hundreds of niches must exist, and different types of bacteria are favored in each of them. The cecal community is therefore characterized by a richness of coexisting bacterial genotypes and phenotypes. The crop must provide a less heterogeneous environment because there is reduced bacterial diversity: the resident community is essentially composed of members of the genus Lactobacillus. From our previous observations, the Lactobacillus population is composed of a limited number of very closely related species (L. crispatus, L. gallinarum, and L. johnsonii) that have homofermentative metabolism and of two disparate species (Lactobacillus reuteri, which is heterofermentative, and Lactobacillus salivarius, which is homofermentative) (15). These species, it must be supposed, are sufficiently different in phenotype to permit their coexistence in this simpler environment. Of these species, only L. crispatus and L. gallinarum produce S-layers, which we have shown in our analysis to be markedly dissimilar between the species (average identity, 64.6%). The production of molecules conferring different cell surface properties, predicted by the primary amino acid sequences of the different S-layers, might allow the two species to evolve sympatrically by the occupation of different crop niches and hence to cohabit the crop.

A striking feature of the results of microbial ecological studies of gut communities is not only the diversity of species that are present, but also the diversity of genotypes (strains) within a species. This has been demonstrated by comparisons of “genetic fingerprints” generated by restriction endonuclease digestion of bacterial genomes followed by agarose gel electrophoresis of one type or another (19, 26, 37). This diversity of strains was again demonstrated in our study, since we detected 17 SmaI genotypes of L. gallinarum among 38 crop isolates. It is difficult to imagine how all of these strains can coexist in the crop unless differential phenotypes exist (16). In fact, although we have investigated in vivo only two of the many strains that we isolated, it appears that each strain differs in the S-protein gene that has favored expression in the crop. This altered phenotype may be sufficient to allow the different strains that we isolated from the birds to occupy different, as yet unidentified, crop niches and hence to coexist in the same host. We do not know the importance of S-proteins to lactobacilli inhabiting the crop, but we observed that both of the L. gallinarum strains that we tested in vivo were able to adhere to the crop epithelium. The recently reported role of a large cell surface protein of L. reuteri in the association of Lactobacillus cells with the forestomach epithelium of mice suggests that S-proteins could have a similar function (43). If this assumption is true, then different S-layer proteins may interact with different receptors on crop epithelial cells, hence providing the opportunity for a variety of L. gallinarum strains to live together without direct competition for attachment sites on the crop epithelium.

It is intriguing that two S-protein genes were detected in each strain of L. gallinarum and that the expression of a single gene was detected in each case. Mutation is considered to be the ultimate source of genetic variation. Therefore, the presence of two similar genes in a genome may reflect gene duplication followed by conservation of the ancestral gene while mutations accrue in the duplicate gene which is transcribed (14, 25, 41). The mutations may result in a gene product that enhances the ecological fitness of the strain in a given environment. This mechanism has been implicated in the evolution of novel morphological features of plants and animals (28); it may also apply to the fitness of L. gallinarum strains for the crop.

Our observations of S-protein genes and their expression by L. gallinarum provide a basis on which more detailed experiments can be planned to study the evolution and ecology of autochthonous gut bacteria. Bacteria have an enormous capacity, because of their rapid generation times and large populations, to evolve and diversify through natural selection. The multitude of ecological niches that exist in the digestive tracts of animals make gut microbiota-host relationships a potentially fertile field of study on the evolution of microbial diversity.