1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements

A colony immunoblotting method has been developed to allow detection of the probiotic Bifidobacterium animalis strain DN-173 010 in human faecal samples. Rabbits were immunized with heat-killed DN-173 010 bacteria resulting in the production of an antiserum highly specific for bacteria belonging to Bif. animalis species. Of the 89 strains representative of 29 different bifidobacterial species tested, only the 15 strains of the Bif. animalis species could be detected with the antiserum. In Western immunoblotting the serum reacts with a protein of 45-kDa apparent molecular weight. None of the bacteria classically encountered in human faecal samples and able to grow on non-selective Columbia blood agar (enterobacteria, Bacteroides or Lactobacillus for instance) reacted with the antiserum. Taking advantage of the high specificity of the antiserum and of the absence of Bif. animalis bacteria in faeces samples of five human volunteers, we demonstrated that strain DN-173 010 survives the intestinal transit. Being based on a combination of semiselective cultivation and colony immunoblotting techniques, the method allowed detection of the Bif. animalis strain even when it represented only one thousandth of the total bifidobacterial population.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements

Understanding the mechanism which sustain a probiotic effect requires analysis of the behaviour of a strain during its transit in the intestinal tract. Survival analysis of a bifidobacterial strain in vivo in humans is still greatly hampered by the lack of specific and quantitative detection methods. During the past 10 years, several methods aiming at tracking bifidobacteria in complex environments have been developed ( Charteris et al. 1997 ; O’Sullivan and Kullen 1998). These methods are based on the use of bacteriological media, more or less selective for bifidobacteria, and traditional identification methods, or the use of DNA probes or molecular typing methods such as ribotyping, pulse field gel electrophoresis, restriction length polymorphism of amplified 16S genes or random amplified polymorphic DNA. The specificity of these methods is variable from the genus to the species or strain level.

Bifidobacterium animalis is a frequently encountered species in fermented milk products ( Biavati et al. 1992 ). A DNA probe specific for Bif. animalis has been described by Mangin et al. (1995) . It was tested only with two Bif. animalis strains and its effectiveness at specifically tracking this bifidobacterial species in the human intestine has not been tested.

In the present study, we developed a method allowing the specific tracking of the Bif. animalis strain DN-173 010 during intestinal transit. This method is based on the combined use of a selective bacteriological media, a species-specific polyclonal serum and a colony immunoblotting protocol.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements

Bacteria and culture media

Bacterial strains used in this study are listed in Table 1. Bifidobacteria were cultivated in trypticase peptone yeast (TPY) broth or MRS pH 6·8 (MRSn) broth (Difco Laboratories, Detroit, MI, USA) complemented with 0·3 g l−1 cysteine hydrochloride monohydrate, under anaerobic conditions (N2/CO2/H2 80/10/10 v/v) or in jars with anaerobiosis generated with AnaeroGen (Oxoid, Basingstoke, UK) at 37 °C. Colonies were obtained under anaerobic conditions at 37 °C using MRSn, Columbia blood (Difco Laboratories), Beerens pH 5·0 ( Beerens 1990), Wilkins Chalgren or Beerens pH 5·5 agar media.

Table 1.  Designation and origin of bifidobacteria strains used in this study and cross-reactivity with anti-DN-173 010 serum under stringent colony-immunoblotting
Name as receivedReferenceOriginImmunodetection
  1. ATCC – American Type Culture Collection, 10801 University Boulevard, Manassas, VA 20110-2209, USA.

  2. CCUG – Culture Collection, University of Göteborg, Department of Clinical Bacteriology, Institute of Clinical Bacteriology, Immunology, and Virology, Guldhedsgatn 10A s-413,46.

  3. CUETM – Collection de l’Unité d’Ecotoxicologie Microbienne, Institut National de la Santé et de la Recherche Médicale, 59651 Villeneuve d’Ascq, France.

  4. DSM=DSMZ – Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Mascheroder Weg 1B, D-38124 Braunschweig, Germany.

  5. NCFB – National Collection of Food Bacteria, AFRC Institute of Food Research, Shinfield, Reading RG2 9AT, UK. (Previously named NCDO.)

  6. NCTC – National Collection of Type Cultures, Central Public Health Laboratory, Colindale Ave., London NW9 5HT, UK.

Bif. adolescentisDN-150 005T = DSM20083 Intestine of adult
CUETM 89/14T = CCUG 18363 Intestine of adult
DN-150 010Human faeces
DN-150 034Human faeces
DN-150 017Human faeces
DN-150 011Human faeces
DN-150 036Human faeces
DN-150 037Human faeces
Bif. angulatumCUETM 89/12T = DSM 20098 Intestine of adult
Bif. animalisCUETM 89/13T = NCFB 2242 Rat faeces
CUETM 92/53Rat faeces+
CUETM 92/54Chicken faeces+
CUETM 92/55Sewage+
CUETM 92/58Rabbit faeces+
CUETM 92/59Sewage+
CUETM 92/60 = ATCC 27674Rabbit faeces+
DN-151 059Human faeces+
DN-173 010Fermented milk+
DN-151 040 = ATCC 27536Chicken faeces+
DN-151 043 = ATCC 27673Sewage+
DN-151 044 = ATCC 27674Rabbit faeces+
Bif. asteroidesCUETM 89/55T = DSM 20089 Hindgut of honeybee
Bif. bifidumCUETM 89/27T = DSM 20082 Human faeces
DN-154 053Human faeces
DN-154 058Human faeces
DN-154 054Human faeces
DN-154 052Human faeces
DN-154 055Human faeces
DN-154 056Human faeces
DN-154 057Human faeces
DN-154 051Human faeces
Bif. boumCUETM 89/22T = DSM 20432 Rumen
Bif. breveCUETM 90/103T = NCFB 2257 Intestine of infant
Bif. catenulatumCUETM 89/96T = CCUG 18366 Human faeces
Bif. choerinumCUETM 89/45T = DSM 20434 Pig faeces
Bif. coryneformeCUETM 89/28T = DSM 20216 Hindgut of honeybee
Bif. cuniculiCUETM 89/30T = DSM 20435 Rabbit faeces
Bif. dentiumCUETM 89/20T = CCUG 18367 Dental carries
DN-170 002Human faeces
DN-170 003Human faeces
Bif. gallicumCUETM 90/72T = DSM 20093 Human faeces
Bif. gallinarumCUETM 89/234T = ATCC 33777 Chicken faeces
Bif. globosumCUETM 89/33T = ATCC 25865 Rumen
Bif. indicumCUETM 89/37T = DSM 20214 Hindgut of honeybee
Bif. infantisCUETM 89/19T = DSM 20088 Human faeces
Bif. lactisDN-166 001T = DSM 10140 Fermented milk+
DN-173 028 = Hansen BB12Fermented milk+
Bif. longumCUETM 89/11T = NCTC 11818 Intestine of adult
DN-164 074Human faeces
DN-164 063Human faeces
DN-164 066Human faeces
DN-164 067Human faeces
DN-164 065Human faeces
DN-164 068Human faeces
DN-164 069Human faeces
DN-164 070Human faeces
DN-164 071Human faeces
DN-164 072Human faeces
Bif. magnumCUETM 89/21T = DSM 20222 Rabbit faeces
Bif. merycicumCUETM 91/25T = RU 915Biavati Rumen
Bif. minimumCUETM 89/24T = DSM 20122 Sewage
Bif. pseudocatenulatumCUETM 89/16T = DSM 20438 Human faeces
DN-168 002Human faeces
DN-168 003Human faeces
DN-168 005Human faeces
DN-168 004Human faeces
DN-168 006Human faeces
DN-168 007Human faeces
DN-168 012Human faeces
CUETM 89/153Human faeces
CUETM 89/241Human faeces
CUETM 93/40Human faeces
CUETM 91/84Human faeces
CUETM 91/99Human faeces
CUETM 89/212Human faeces
CUETM 91/92Human faeces
CUETM 89/98Human faeces
CUETM 92/15Human faeces
CUETM 93/52Human faeces
CUETM 89/173Human faeces
CUETM 90/136Human faeces
CUETM 89/29Human faeces
Bif. pseudolongumCUETM 89/18T = DSM 20099 Pig faeces
Bif. pullorumCUETM 89/17T = DSM 20433 Chicken faeces
Bif. ruminantiumCUETM 91/24T = RU 687Biavati Rumen
Bif. sp.DN-163 040 = Wisby 420Fermented milk+
Bif. subtileCUETM 89/35T = DSM 20096 Sewage
Bif. suisCUETM 89/42T = SU 859Biavati Pig faeces
Bif. thermophilumCUETM 97/02T = MB1Biavati Pig faeces

DNA–DNA hybridization

Experiments were carried out on 12 strains selected from the numerical analysis.

DNA was prepared with the simultaneous use of achromopeptidase (Sigma, Saint Quentin, Fallavier, France; 5000 U 500 mg bacteria−1) ( Barsotti et al. 1988 ) and lysozyme (400 000 U 500 mg bacteria−1) as lytic enzymes. DNA extraction was carried out according to Marmur’s method (1961).

Degree of DNA–DNA binding was determined quantitatively by spectrophotometry from renaturation rates in accordance with a modification of the method of De Ley et al. (1970) . The temperature of renaturation was Tm − 25 °C. DNA–DNA relatedness values were calculated after incubation of 21 and 24 min, following removal in calculation of the first 3 min of renaturation.


Antibodies were raised against Bif. animalis strain DN-173 010, in New Zealand white rabbits. Bacteria were cultivated without agitation for 16 h at 37 °C in MRSn broth supplemented with 300 mg l−1 cysteine, centrifuged at 5000 g and kept at − 80 °C. One milligram of thawed bacterial pellet, incubated 20 min at 100 °C, was resuspended in 1 ml of sterile 0·9% NaCl solution and emulsified in 1 ml of Freund’s complete adjuvant. On day 0 the bacterial suspension was injected intramuscularly. On days 21, 42 and 63, bacteria were prepared as previously but emulsified in Freund’s incomplete adjuvant were injected. Reactivity of sera was determined by the dot blot technique. The hyperimmunization was obtained after the fourth injection. The maximum and constant production of antibodies was obtained by further repetitive injections at intervals of 3 weeks. In general, rabbits were bled 10 d after the fourth injection and sera decomplemented by 30-min incubation at 56 °C, a procedure which was used for all experiments. Serum from the last bleeding was used to set up the colony immunoblotting method.

Immunodetection techniques

Colony immunoblotting. Colonies of bifidobacteria obtained on agar plates were transferred to 0·45-µm nitrocellulose filters (Protran BA 85, Schleicher and Schuell, Dassel, Germany). After transfer of the colonies the nitrocellulose filter was fixed for 30 min at 80 °C, rehydrated and incubated for 1 h with agitation in freshly prepared blocking solution (PBS, Gibco; 0·2% Tween 20, Sigma; 5% skim milk powder). The filter was then incubated under the same conditions in anti-DN-173 010 serum diluted 1/64 000 in the blocking solution. The filter was washed three times for 5 min with blocking solution, incubated for 1 h in blocking solution containing 0·1 µg ml−1 (around 0·015 units ml−1) of horse radish peroxidase-conjugated protein A (HRP-protein A, Sigma), washed twice for 5 min in washing solution (PBS; 0·2% Tween 20) and once for 5 min in PBS. Colonies reacting with antibodies of the serum were detected using the diaminobenzidine colourimetric detection system (DAB, Sigma). Control filters treated under the same conditions except for the omission of the serum were included in each experiment.

When needed, transfer and precise location of all colonies to the nitrocellulose filter was checked prior to the immunodetection, using the Ponceau S (Sigma) detection method and following manufacturer’s instructions.

Western blotting. Bacteria from a 10-ml dense culture of DN-173 010 were harvested by centrifugation, washed twice in PBS, resuspended in 1 ml sample buffer (Tris-HCl 0·06 mol l−1; SDS 3%; glycerol 10%; β-mercaptoethanol 0·71 mol l−1), incubated for 18 h at 4 °C and pulse-sonicated three times for 60 s at 4 °C. The whole cells and cellular debris were pelleted and discarded by centrifugation. Supernatant (5–15 µl) was mixed with 5 µl of a 0·3% bromophenol blue solution and heated for 5 min at 95 °C before loading on a 10% acrylamide gel. After SDS-PAGE separation, protein bands were transferred to a nitrocellulose membrane. The transfer efficiency was controlled by Ponceau S staining. Proteins reacting with the anti-DN-173 010 serum were detected as described for colony immunoblotting using a 1/6000 dilution of the serum.

Analysis of human faecal samples

Addition of strain DN-173 010 to human faeces. Human faeces were analysed after 10-fold dilutions in cysteinated quarter strength Ringer diluant ( Neut et al. 1989 ) and plating on non-selective Columbia blood agar and selective Beerens’ agar at pH 5·0 and 5·5. Dilutions were spread on plates directly and after mixture with dilutions of strain DN-173 010. All plates were incubated under anaerobic conditions at 37 °C for 1 week before counts and the obtained colonies were transferred to filters for immunoblotting.

Volunteers. Five healthy female volunteers (mean age 33, range 20–48 years) without any history of bowel disease and without antibiotherapy for at least 3 months ingested three times daily 125 g of commercial BIO®-fermented milk for 1 week. A first faecal sample was obtained before starting ingestion, a second after 1 week of ingestion.

Analysis of BIO®-fermented milk. Three different batches of commercial BIO®-fermented milk products were used. The products were stored at 4 °C. Volunteers 1 and 2 ingested products from the first batch, respectively, between days 12 and 18 and between days 15 and 21 after the date of production of the fermented milk. Volunteers 3 and 4 ingested products from the second batch, respectively, between days 10 and 16 and between days 11 and 17 after the date of production of the fermented milk. Volunteer 5 ingested products from the third batch between days 10 and 16 after the date of production of the fermented milk. Enumeration of strain DN-173 010 present in BIO®-fermented milk was performed in MRSn agar complemented with 0·3 g l−1 cysteine and 0·5 mg l−1 dicloxacilline after a 5-d incubation of the plates under anaerobic conditions. The three batches contained, respectively, 7·98 log (cfu) g−1, 8·10 log (cfu) g−1 and 8·04 log (cfu) g−1 of strain DN-173 010 (mean value of four counts, two performed on day 8 and two performed on day 22 after the date of production).

Sampling. Fresh voided faeces were collected in a sterile 400-ml plastic container; the atmosphere was rendered anaerobic by addition of a humidified Anaerocult® (Merck, Darmstadt, Germany). The sample was immediately transported to the laboratory and analysed within 4 h.

Bacteriological analysis. Growth of faecal bifidobacteria is usually checked on the selective Beerens’ medium ( Beerens 1990). But the pure strain DN-173 010 showed only very sparse growth on this medium giving pinpoint colonies after growth for 4 d under anaerobic conditions. However, when Beerens’ medium was adjusted to pH 5·5 (and not 5·0 as described by the author) growth as abundant as at neutral pH was observed. Therefore, for the analysis of faeces we chose Beerens’ medium at pH 5·0 for outnumbering bifidobacteria, but added another plate with Beerens’ medium at pH 5·5 allowing growth of strain DN-173 010. The second medium is surely less selective for bifidobacteria and might allow the growth of other micro-organisms. Counts on these plates do not reflect counts of bifidobacteria only.

The faecal sample (around 1 g) was introduced into the first preweighed tube of the dilution series containing 9 ml of quarter-strength cysteinated Ringer solution. This tube was reweighed after introduction of the sample for the determination of the exact weight. Then eight further 10-fold dilutions were made in the same diluant. Each dilution (0·1 ml) was spread on one plate containing non-selective modified Columbia blood agar, one plate containing Beerens’ selective medium at pH 5·0 and two plates containing Beerens’ selective medium at pH 5·5. All plates were incubated for 5 d at 37 °C under anaerobic conditions. For each faecal sample and media the plate showing 15–150 colonies was used for calculations.

The colonies on Columbia blood were outnumbered to obtain counts of the total cultivable anaerobe flora. Colonies on Beerens’ plates were transferred to nitrocellulose filters and subjected to immunodetection to obtain counts of Bif. animalis. Ten colonies of different sizes or aspects, giving a positive signal or not by the colony immunodetection method were subcultured for identification.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements

Specificity and sensitivity of the anti-DN-173 010 serum

Tests with pure bacterial cultures. Strain DN-173 010 was used to raise antibodies in rabbits. The serological cross reactivity of 89 strains covering 29 species of the Bifidobacterium genus was tested using a colony immunoblotting method ( Table 1). Bacterial spots were obtained by inoculating a single agar plate with 5-µl drops of pure bacterial suspensions for each strain. In addition to strain DN-173 010, only Bif. animalis strains, Bif. lactis strains and one Bifidobacterium sp. strain showed cross-reactivity with the immune serum ( Table 1). In order to clarify the specificity of the immune serum, both the DN-173 010 strain and the Bifidobacterium. sp. strain Wisby 420 were identified at the species level using DNA–DNA hybridization in solution ( Table 2). Both strains were found to belong to the Bif. animalis species (respectively 95 and 90% homology with the Bif. animalis type strain DNA). DNA from the two Bif. lactis strains was also included in the DNA–DNA hybridization studies and was found to be highly homologous with Bif. animalis DNA (87 and 86% homology). Bif. lactis has recently been proposed as a new species on the basis of DNA–DNA hybridization studies performed on nitrocellulose filters, a less common method for species determination ( Meile et al. 1997 ). Our results show that the two Bif. lactis strains belong to the Bif. animalis species rather than to a new bifidobacterial species. Identification at the species level of all the other tested strains was confirmed by phenotypic characteristics and checked by DNA–DNA hybridization (except for Bif. bifidum, species with a typical phenotype).

Table 2.  Percentage of homology of DN-173 010, Bifidobacterium sp. Wisby 420, and two strains of Bif. lactis with the Bif. animalis type strain
StrainBif. animalis CUETM 89/13T (%)
DN-173 01095
Bif. sp. Wisby 42090
Bif. lactis Hansen BB1286
Bif. lactis DSM 10140T87

Results presented in Table 1 were obtained using an optimized protocol as described in Materials and methods. Under these rather stringent optimized conditions, the Bif. animalis type strain could hardly be detected. Using less stringent conditions, i.e. lower dilutions of the serum, bovine serum albumin as a blocking agent instead of milk, goat antirabbit immunoglobulins conjugated to HRP instead of HRP-protein A or lower dilutions of HRP-protein A, all Bif. lactis and Bif. animalis strains tested, including the Bif. animalis type strain, could be detected. Under these nonoptimized conditions the Bif. adolescentis, Bif. breve, Bif. dentium and Bif. infantis type strains, and all the Bif. bifidum strains tested, including the type strain were still not detected, but the Bif. catenulatum and Bif. thermophilum type strains and all the Bif. pseudocatenulatum strains tested except the type strain were detected. Detection of these strains was shown to be due to direct binding of the goat antirabbit antibodies or of the HRP-protein A to the bacteria (data not shown).

The antiserum was also tested using different immunological methods. Using ELISA, immunoelectrophoresis or Ouchterlony immunodiffusion, cross-reactivity of the serum with the Bif. animalis type strain could be observed (data not shown). Under Western immunoblotting conditions, in total protein extracts of Bif. animalis and Bif. lactis strains, a protein with a molecular weight of about 45 kDa was reacting with the anti-DN-173 010 serum. The major cross-reacting protein of the Bif. animalis type strain is slightly smaller than that of the DN-173 010 and the Bif. lactis type strain. To a lesser extent, the serum also reacts with a protein of about 95 kDa common to all bifidobacterial strains tested ( Fig. 1).


Figure 1. Western blot of total extracts of bifidobacterial strains. Mw– molecular weight standards, lane 1 –Bifidobacterium animalisT, lane 2 – DN-173 010, lane 3 –Bif. lactisT, lane 4 –Bif. globosumT, and lane 5 –Bif. pseudocatenulatumT. Detection by anti-DN-173 010 serum as described in Material and methods

Download figure to PowerPoint

From both the immunoblotting data and results of DNA–DNA hybridization, we conclude that the serum raised against strain DN-173 010 allows specific detection of Bif. animalis and that among Bif. animalis strains, DN-173 010 appears closer to the so-called Bif. lactis strains than to the Bif. animalis type strain.

Tests with strain DN-173 010 added to faecal samples. Since the serum and the colony immunoblotting protocol were designed to allow specific detection of Bif. animalis DN-173 010 in faecal samples, we checked that no bacterial strains classically present in faecal samples and able to grow on the used bacteriological media could react with the serum. Colonies obtained by plating dilutions of human faecal samples with or without addition of strain DN-173 010 prior to plating were analysed by immunoblotting. No cross-reactivity was observed with bacteria of the normal faecal flora even when non-selective media was used. Cross-reacting colonies were detected only when strain DN-173 010 had been added to the faecal sample prior to plating, confirming the very narrow specificity of the serum for bacteria of the Bif. animalis species ( Table 3).

Table 3.  Detection by colony immunoblotting of strain DN-173 010 added to faecal samples
Plated suspension *Total colonies Positive colonies
  • *

    Dilutions and volumes were determined to obtain between 10 and 100 colonies per plate.

  • Colonies enumerated on filters after Ponceau S staining and before immunodetection.

  • Colonies reacting with the anti-DN-173 010 serum under stringent colony immunoblotting.

  • §

    10-fold dilution of the DN-173 010 suspension in order to divide by 10 the number of DN-173 010 colonies, keeping constant the number of colonies from the faeces sample.

Faeces sample330
DN-173 0104646
Faeces sample + DN-173 01013363
Faeces sample + DN-173 010 (1/10) §677

A single colony reacting with the serum could be detected on plates containing a total of about 1000 colonies (data not shown) suggesting that Bif. animalis DN-173 010 could be detected in faecal samples even if it represents only one thousandth of the selected bacterial population.

Volunteers experiments

The colony immunodetection tool was shown to be a suitable tool for qualitative and quantitative detection of DN-173 010 in faecal samples of human volunteers who had ingested fermented milk products containing that strain. Faecal samples of five human volunteers were analysed before and after a 7-d period of ingestion of DN-173 010-containing fermented milk ( Fig. 2). About 10·5 log (cfu) of DN-173 010 were ingested daily. This amount is identical for each volunteer and stable over 7 d.


Figure 2. Enumeration of the bifidobacterial and the Bifidobacterium animalis population in faecal samples of five human volunteers before and after 1 week’s ingestion of BIO®-fermented milk. The bifidobacterial population was enumerated on Beerens’ agar pH 5·0 and thus does not include Bif. animalis strain DN-173 010. The Bif. animalis population (positive colonies) was enumerated using Beerens’ agar pH 5·5 and the colony immunoblotting method based on anti-DN-173 010 serum

Download figure to PowerPoint

The bacterial population enumerated on Beerens’ medium at pH 5·0 corresponded to the major bifidobacterial population, probably excluding the DN-173 010 population since this latter strain, if present, gave colonies which were too small to be detected. Bacterial counts varied between the different volunteers, and ranged from 5·7 to 9·4 log (cfu) g−1 on day 0 and from 5·9 to 10·1 log (cfu) g−1 on day 7. It was either unchanged or slightly increased (1·26 log maximum) on day 7 compared with day 0.

No colonies of Bif. animalis were detected in any of the faecal samples before ingestion of the fermented milk, thus confirming that Bif. animalis is only seldomly encountered in the human intestinal flora in the absence of probiotic ingestion (to be published elsewhere). After 7 d of ingestion of the fermented milk, bacteria reacting with the immune serum could be detected in faecal samples of all five volunteers ( Fig. 3). For each volunteer 10 colonies from Beerens’ medium at pH 5·5, reacting with the serum or not, were subcultured and identified by classical bacteriological methods and DNA–DNA hybridization. Each cross-reactive colony that could be subcultured was identified as Bif. animalis. The non-cross-reacting colonies were identified as Bif. longum, Bif. adolescentis, Bif. pseudolongum, Bif. bifidum or Bif. pseudocatenulatum. Since no Bif. animalis strain was detectable in the faecal content of any of the volunteers before ingestion of the Bif. animalis DN-173 010 fermented milk and since no other fermented dairy products were ingested during the 7 d of the study, we concluded that the probiotic strain DN-173 010 could be detected at high levels in the faecal samples of the volunteers.


Figure 3. Detection by colony immunoblotting of strain DN-173 010 in human faecal samples after ingestion of fermented milk. Immunoblotting was performed with anti-DN-173 010 serum as described in Materials and methods. (a) Ponceau S staining showing total colonies, and (b) immunoblotting of the same filters showing positive colonies only

Download figure to PowerPoint

The faecal population of DN-173 010 after 7 d of ingestion of the strain was about 9 log (cfu) g−1 in four volunteers and 8 log (cfu) g−1 for one volunteer. Interestingly this latter volunteer exhibiting the lowest population of DN-173 010 corresponded with the volunteer with the lowest overall total bifidobacterial faecal population.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements

The colony immunoblotting method described in this report allows simple and quantitative detection of Bif. animalis strain DN-173 010 in human faecal samples.

This method is based on the use of a rabbit polyclonal serum directed against strain DN-173 010 and reacting with a protein of about 45 kDa. This protein is probably present at the surface of the bacterial cells since heat-killed bacteria were used to raise the antibodies and in the colony hybridization procedure no lysis step is necessary to detect bacteria. The protein is also highly immunogenic and/or abundant since it can still be detected with a 1 : 64 000 dilution of the serum. In addition it corresponds with a Bif. animalis species-specific epitope since all Bif. animalis strains were tested and no strains of the 28 other bifidobacterial species reacted with the serum. Interestingly, among species of the Bifidobacterium genus, Bif. animalis is probably not the only one expressing highly immunogenic and/or abundant species-specific surface epitopes. Using one Bif. breve strain and one Bif. bifidum strain, Bouquelet et al. also obtained rabbit polyclonal sera allowing specific detection of these two species, respectively (personal communication). Presence of species-specific epitopes in bifidobacterial species could be a consequence of the rather distant phylogenetic distance separating the different species of the Bifidobacterium genus.

The colony immunoblotting method relies on a colony growth step, a step often considered critical due to the selectivity of the bacteriological media and incubation conditions used. Use of this step might lead to an underestimation of the Bif. animalis population, however, it presents the double advantage of allowing detection of only viable bacteria, and increasing the sensitivity of the detection by acting as an enrichment step. Use of the Bif. animalis-specific serum for direct immunofluorescence labelling of bacteria in fixed faeces samples and epifluorescence microscopy could represent an alternative method avoiding the colony growth step.

The serum directed against DN-173 010 is species specific and not strain specific. Thus it can be used to track other Bif. animalis strains i.e. the strains most frequently found in commercial fermented dairy products ( Biavati et al. 1992 ). In addition to its use for the detection of bacterial strains in complex microbial environments, it represents a valuable complementary tool for strain identification at the species level.

Using DNA–DNA hybridization in solution, we clearly demonstrated that strain DN-173 010, strain Wisby Bifidobacterium sp. 420 and two strains originally designated as members of the Bif. lactis species, DSM 10 140 and Hansen BB12 are indeed strains of the Bif. animalis species. These strains could represent a group of closely related strains in the Bif. animalis species, more closely related to each other than to the Bif. animalis type strain. The major protein detected with the anti-DN-173 010 serum has the same apparent molecular weight in strain DN-173 010 and strain DSM 10 140, and is slightly bigger than that of the Bif. animalis type strain. In addition, under colony immunoblotting conditions, cross-reactivity of the Bif. animalis type strain is weaker than that of the other Bif. animalis strains. Finally, oligonucleotide probes allowing detection by PCR of strain Bifidobacterium sp. Wisby 420 have been described ( Kok et al. 1996 ). These probes are not strain specific since they detect the Bif. animalis strains DN-173 010, DSM 10 140, ATCC 27 536 and ATCC 27 674 (unpublished results) but they do not detect the Bif. animalis type strain. Apart from these differences, the type strain of the Bif. animalis species is not known as phenotypically or genotypically different from other Bif. animalis strains such as other type strains can be.

The analysis performed with a group of five human volunteers demonstrates the validity of the colony immunoblotting method to track strain DN-173 010. It also demonstrates that strain DN-173 010 survives human intestinal transit. Its survival kinetics in the different compartments of the digestive tract and the influence of different parameters on these kinetics remains to be analysed.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements

We thank J.-M. Antoine for helpful discussions, F. Rassat and S. Doat for yoghurt preparation and analysis, C. Aubert-Jacquin for assistance in yoghurt supply and M.-C. Degivry for technical assistance.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  • Barsotti, O., Morrier, J.J., Freney, J., Renaud, F., Benay, G., Decoret, D., Dumont, J. (1988) Achromopeptidase for rapid lysis of oral anaerobic Gram-positive rods. Oral Microbiological Immunology 3, 86 88.
  • Beerens, H. (1990) An elective and selective isolation medium for Bifidobacterium spp. Letters in Applied Microbiology 11, 155 157.
  • Biavati, B., Mattarelli, P., Crociani, F. (1992) Identification of bifidobacteria from fermented milk products. Microbiologica 15, 7 14.
  • Charteris, W.P., Kelly, P.M., Morelli, L., Collins, J.K. (1997) Selective detection, enumeration and identification of potentially probiotic Lactobacillus and Bifidobacterium species in mixed bacterial populations. International Journal of Food Microbiology 35, 1 27.
  • De Ley, J., Cattoir, H., Reynaerts, A. (1970) The quantitative measurement of DNA hybridization from renaturation rates. European Biochemistry 12, 133 142.
  • Kok, R., De Waal, A., Schut, F., Welling, G.W., Weenk, G., Hellingwerf, K.J. (1996) Specific detection and analysis of a probiotic Bifidobacterium strain in infant feces. Applied and Environmental Microbiology 62, 3668 3672.
  • Mangin, I., Bourget, N., Simonet, J.M., Decaris, B. (1995) Selection of species-specific DNA probes which detect strain restriction polymorphism in four Bifidobacterium species. Research in Microbiology 146, 59 71.
  • Marmur, J. (1961) A procedure for isolation of deoxyribonucleic acid from microorganisms. Journal of Molecular Biology 3, 208 218.
  • Meile, L., Ludwig, W., Rueger, U., Gut, C., Kaufmann, P., Dasen, G., Wenger, S., Teuber, M. (1997) Bifidobacterium lactis sp. nov., a moderately oxygen tolerant species isolated from fermented milk . Systematic and Applied Microbiology 20, 57 64.
  • Neut, C., Colombel, J.F., Guillemot, F., Cortot, A., Quandalle, P., Ribet, M., Romond, C., Paris, J.C. (1989) Impaired bacterial flora in human excluded colon. Gut 30, 1094 1098.
  • O’Sullivan, D.J. & Kullen, M.J. (1998) Tracking of probiotic Bifidobacteria in the intestine. International Dairy Journal 8, 513 525.