- Email: [email protected]
Monitoring and Survival of Lactobacillus paracasei LTH 2579 in Food and the Human Intestinal Tract
SYSTEIVIATIC AND _htt--,---p:_//ww_w_ .ur_ba_nf_isc_h_er._de--'.,/jo_u_rna_ls_/s_am_ _ _ _ _ _ _ _ _ _ _ _ APPUED MICROBIOLOGY System. Appl. Microbiol. 23, 260-266 (2000)
© Urban & Fischer Verlag
Monitoring and Survival of Lactobacillus paracasei LTH 2579 in Food and the Human Intestinal Tract CHRISTOPH BUNTE, CHRISTIAN HERTEL, and WALTER P. HAMMES
Institut fur Lebensmitteltechnologie, Universitat Hohenheim, Stuttgart, Germany Received April 3, 2000
Summary A PCR-based detection system specific for Lactobacillus paracasei LTH 25 79 was developed and applied to follow the fate of the strain in complex ecosystems. This strain was isolated from fruit mash and was characterised as being highly resistant to low pH and bile at concentrations as they occur in the human digestive tract. The application of the subtraction hybridisation technique permitted to identify a 235 bp chromosomal DNA fragment of strain LTH 2579. Based on this target sequence a specific PCR system was developed and combined with the species-specific PCR system for L. paracasei. This combination of PCR based detection systems was successfully applied to monitor L. paracasei LTH 2579 in fermented sausages which were inoculated with this strain (2.0 x 10 7 CFU/g) together with the strongly competitive L. sakei LTH 681 (1.0 x 10 6 CFU/g). At the time of consumption of the sausages the respective counts were 1.8 x 10 7 and 1.4 x 10 8 CFU/g. After consumption of the sausages by three volunteers L. paracasei LTH 2579 was recovered from fecal samples. The counts determined for the strain ranged between 1.2 x 10 7 and 1.5 x 10 8 CFU/g of feces. The fortuitous lactobacilli constituted a share of 5-12% of the lactobacilli in the fecal flora. Key words: strain-specific PCR - species-specific PCR - subtraction hybridisation - Lactobacillus para-
casei - Lactobacillus sakei - intestinal flora - fermented food
Introduction To monitor the fate of a bacterial strain implanted into complex ecosystems, unique phenotypic or genotypic markers are needed for its reliable and specific detection. The phenotypic approach, e.g. the application of serological techniques (HENRIKSEN, 1978), depends on the expression of the phenotypic marker and may thus be affected by environmental factors. On the other hand, genotypic detection methods with taxonomic resolution to the strain level (VANDAMME et al., 1996) have been proven more useful, e.g. restriction fragment length polymorphism (RFLP) and random amplified polymorphic DNA (RAPD-PCR). A further approach is the detection of strain-specific DNA sequences by applying hybridisation or PCR techniques with specific probes or primers. These unique sequences can be derived, for example, from RAPD fragments, as it was recently described for the detection of Lactobacillus rhamnosus Lc 1/3 (TILSALA-TIMISJARVI and ALATOSSAVA, 1998) or may be obtained by the subtraction hybridisation technique (BJOURSEN and COOPER, 1988; MATHESON et al., 1997). The applicability of this method to develop a strain-spe0723-2020/00/23/02-260 $ 12.00/0
cific PCR-based detection system has recently been demonstrated by WASSILL et al. (1998). It was the purpose of our investigations to make use of this technique and to develop a specific PCR system for Lactobacillus paracasei LTH 2579. This strain was isolated from fermenting fruit mash and its properties were compared with those of intestinal lactobacilli (HAMMES et al., 1997). It had been shown that strain LTH 2579 is resistant to low pH and bile at concentrations as they occur in the human digestive tract. The strain also adhered to Caco 2-cell cultures. It was our aim to establish a model for monitoring bacterial strains such as starter or probiotic organisms in both fermented food and fecal samples.
Materials and Methods Bacterial strains. growth conditions. and physiological characterisation
The strains used in this study were Lactobacillus sakei DSM 20017T , Lactobacillus curvatus DSM 20019 T , L. sakei LTH 681
Monitoring of L. paracasei LTH 2579 and L. curvatus LTH 684 formerly referred to as strains Ls12 and Lc3 (HAMMES et al., 1990), respectively, and those strains compiled in Table 1. Lactobacilli were grown in MRS medium (DE MAN et al., 1960) at 30°C. For selective detection of intestinallactobacilli, LAMVAB medium (HARTEMINK et al., 1997) was used, and incubations were performed at 37°C under anaerobic conditions (Anaerocult A, Merck). Physiological characterisation of the lactobacilli was performed using the API 50 CH test system (BioMerieux) together with the API CHL medium.
Production and microbial analysis of fermented sausages For the production of 1 kg sausage mixture, 0.90 kg frozen pork meat and 0.10 kg frozen pork belly (40 % lean) were cut and mixed at ambient temperature. Glucose (2 g), nitrite curing salt (26 g), and spice preparation (8 g) were added. Lyophilised starter organisms were suspended in 50 ml water and added to the meat mixture to obtain final cell counts (CFU/g) of 1.0 x 10 6 of L. sakei LTH 681, 8.0 x 10 6 of Staphylococcus carnosus LTH 3127, 1.5 x 10 6 of Kocuria varians LTH 2864, and 2.0 x 10 7 of L. paracasei LTH 2579. Sausages of ca. 68 g were filled in cellulose casings of 20 mm diameter and incubated in ripening chambers (Siidtronic) for 122.5 h. The following ripening conditions were applied: 24 h at 24°C (90-92 % r.H.), 2 h at 24-28 °C (50% r.H.), 30 min at 24-28 °C (smoking), and 96 h at 14-15 °C (75-80% r.H.) . The loss of weight during ripening was 25%. For microbial analysis samples of 10 g of sausage were homogenised with the aid of a stomacher in 90 ml of dilution medium containing per I: 8.5 g NaCI, 1.0 g peptone. Microbial counts of LAB were determined by surface plating on MRS medium supplemented with 0.025 gIl bromocresol green. Microbial analysis of fecal samples Fecal samples were taken from 3 male subjects at 1 day before (sample 1) and on the third day (sample 2) of consumption of 100 glday (subjects 2 and 3) or 200 glday (subject 1). Samples of ca. 0.2 g were transferred to pre-weighed tubes containing 2 ml of a pre-reduced cryoprotective broth (CROWTHER, 1971) which was supplemented with 0.25 gIl cysteine-HCI and 0.15 mgll resazurine. The samples were homogenised, and immediately frozen and stored below -20°C. Cell counts were determined within 3 weeks by decimal dilute in a pre-reduced dilution medium (8.5 gil NaCl, 1.0 gil peptone, and 0.1 gil cysteine) and plating on LAMVAB medium. All anaerobic manipulations were carried out in an anaerobic chamber under an atmosphere of 85% N z, 10% CO 2 , 5% H z. Plates were incubated for 3 days at 37°C under anaerobic conditions. Viable cell counts were expressed as CFU per gram wet weight feces. For counting of L. paracasei LTH 2579, one colony of each colony type was investigated by application of the strain-specific as well as the L. paracasei-specific PCR detection system. The colonies of the type yielding a double positive result were counted, a number of 20 was randomly picked and also subjected to PCR analysis to confirm the homogeneity of the colony type. The cell count of L. paracasei LTH 2579 was finally determined applying the hypergeometric distribution with a level of significance a =5% (SAHAI and KURSHID, 1996). DNA isolation and preparation of crude nucleic acid extracts For isolation of the genomic DNA used as template for PCR, the High Pure PCR Template Preparation Kit (Roche Diagnostics GmbH) was applied. Chromosomal and plasmid DNA of lactobacilli were isolated as described previously (SCHMIDT et al., 1999; CAVADINI et al., 1996). Crude nucleic acid extracts were prepared by resuspending a colony in 40 fli of bidistilled water in a reaction tube. The suspension was heated for 10 min at 98°C and centrifuged at 18,OOOx g for 2 min. An aliquot of the supernatant was subjected to PCR.
261
Design of specific primers To construct the primer specific for Lactobacillus paracasei, 16S rRNA sequences of Lactobacillus paracasei subsp. paracasei, L. paracasei subsp. tolerans, L. casei, L. rhamnosus, and L. zeae were aligned and compared to obtain specific target sites. The EMBL accession numbers of the sequences are D79212, D16550, D16551, D16552, and D86516, respectively. For verification of the specificity, the Probe Match tool of the ARB program package (SPRINGER et al., 1992) was used together with a database containing all deposited complete primary structures of 16S rRNA. For construction of the primer specific for L. sakei, the site within the 23S rDNA which was used previously to design the probe Lbs (HERTEL et al., 1991) was chosen as target site for the primer. To increase the number of mismatches and to place two of three mismatches at the 3' region of the primer, the target site of the primer was shifted 13 bases to the 5' end of the 23S rRNA. For construction of specific primers for L. paracasei LTH 2579 the subtraction hybridisation technique was used as described by WASSIll et al. (1998). The genomic DNA of strain LTH 2579 served as target DNA and that of the L. paracasei strains LTH 2570, LTH 2552, LTH 2553, LTH 2567, and LTH 2600 as subtracter DNA. One DNA fragment of 235 bp was isolated and sequenced applying the dideoxynucleotide chain termination method and universal primers. Based on the sequence specific primers were designed for detection of the strain L. paracasei LTH 2579. Specific and universal polymerase chain reactions All PCR amplifications were carried out on a GeneAmp 2400 Thermocycler (Perkin Elmer) in total volumes of 25 fll. The reaction mixture contained 0.2 mM of each deoxynucleotide triphosphate, 0.5 flM of each primer, PCR buffer (1.5 mM MgCI2' 50 mM KCI, and 10 mM Tris-HCI, pH 9.0), 1.5 U Taq polymerase (Amersham Pharmacia Biotech), and 7 fli of the nucleic acid extract or 50 ng genomic DNA. The following amplification profiles were used: i) for specific detection of L. paracasei: 1 cycle of 2 min at 94°C; 32 cycles of 1 min at 94 °C, 1 min at 49°C, and 30 sec at 72 °C; 1 cycle of 5 min at 72 °C; ii) for specific detection of L. sakei: 1 cycle of 2 min at 94°C; 35 cycles of 1 min at 94 °C, 1 min at 57°C, and 20 sec at 72 °C; 1 cycle of 5 min at 72 °C; iii) for the control reaction with universal primers 616VII and lOOK : 1 cycle of 2 min at 94°C; 32 cycles of 30 sec at 94 °C, 45 sec at 51°C, and 90 sec at 72 °C; 1 cycle of 5 min at 72 0C. For specific detection of the strain L. paracasei LTH 2579 the reaction mix was modified as follows: 0.1 mM of each deoxynucleotide triphosphate, 2.5 U Taq polymerase (Amersham Pharmacia Biotech), and 12 fli of the nucleic acid extract. The amplification profile was the following: 1 cycle of 2 min at 94°C; 35 cycles of 1 min at 94 °C, 1 min at 68°C, and 30 sec at 72 °C; 1 cycle of 5 min at 72 0C. The total contents of the PCR reaction vials were subjected to agarose gel electrophoresis and visualised by staining with ethidium bromide. RAPD polymerase chain reaction RAPD-PCR was performed using primer OPL-02 (TORRIANI et al., 1996). The reaction mix was modified by using 0.2 flM of each primer and 2.5 U Taq polymerase. The amplification profile was the following: 1 cycle of 2 min at 94°C, 45 cycles of 1 min at 94 °C, 1 min at 34°C and 1 min at 72 °C, 1 cycle of 5 min at 72 0C. 10 fli of the content of each PCR reaction vial was subjected to agarose gel electrophoresis and visualised by staining with ethidium bromide. Southern hybridisation Genomic DNA (1 fig) of strain LTH 2579 was digested with the restriction enzymes EcoRI, HindIII, or Pst!. Plasmid DNA
262
C. BUNTE et al.
and the genomic restriction fragments were separated by agarose gel electrophoresis and transferred to a nylon membrane (Quiabrane, Qiagen) using a vacuum blot apparatus (Amersham Pharmacia Biotech) following the protocol of the supplier. After blotting the DNA was immobilised by incubating at 80°C for 30 min. For hybridisation the digoxygenin-labelled probe 234L was constructed using a PCR DIG Probe Synthesis Kit (Roche Diagnostics GmbH), primer pair 2579f / 2579r and genomic DNA of strain L. paracasei LTH 2579 following the instructions of the supplier. Hybridisation (over night) and stringent washing were carried out at 64°C. Detection was performed using the Colorimetric DIG detection kit (Roche Diagnostics GmbH) according to the instruction of the supplier.
Computer analysis The DNA sequences were analysed with the aid of the software DNASIS v2.5 (Hitachi Software Engineering Co.). Amino acid sequences were compared to the Swissprot database (Swiss
Institute of Bioinformatics, Geneva) and to the Protein information resource (National Biomedical Research Foundation, Washington) using BLAST.
Results Development of species-specific detection systems To develop a PCR-based system for specific detection of L. paracasei, the 16S rDNA was chosen as target molecule. Comparative analysis of the 16S rRNA sequences revealed a region that was used as target site for the species-specific primer Lpacaf (Tab. 2). The Computer-based validation of the specificity showed that all potential primer target hybrids contained at least one weighted mismatch. Combination of this primer with the
Table 1. Specificity of the PCR with primer pairs Lpacaf /97 K and 2579f / 2579r. Strain
Species
PCR with primer pair
Source
Lpacaf /97 K
L. L. L. L. L.
casei zeae rhamnosus paracasei subsp. tolerans paracasei subsp. paracasei
L. casei
Lactobacillus spec. a
DSM 20011 T DSM 20178 T DSM 20021 T DSM 20258 T DSM 5622 T DSM20008 DSM20020 DSM20207 DSM 20312 LTH2579 (2 strains) (2 strains) (7 strains) (1 strain) (8 strains) (1 strain) (5 strains) (4 strains) (48 strains)
2579f / 2579r
cheese corn steep liquor milk
+ + + + + + + + + + + + + + +
milk saliva milk feces apple mash sauerkraut cherry mash pear mash plum mash sauerkraut beer unknown feces a feces a
+ +
isolated on LAMVAB agar from 22 subjects
Table 2. Specific and universal primers used in the PCR systems. Primer
Sequence (5'-3')
Specificity
Position
Lpacaf 97K Lsakf 383R 2579f 2579r 616VII lOOK
CCGAGATTCAACATGG CTGCTGCCTCCCGTA CCCAATACTTTTAATCGA GTTCCGCCGTACTCA GATTTAGTCTGCACACGC CCTTCACTGTCATCATCC AGAGTTTGATYMTGGCTCAG GGTTACCTTGTTACGACTT
L. paracasei universal L. sakei group-specific L. paracasei LTH 2579 L. paracasei LTH 2579 universal universal
88 a,b 357 a,b 128 a,' 383 a,c 2d 235 d 7 a,b 1491 a,b
according to E.coli numbering 16S rDNA c 23S rDNA d 235 bp DNA fragment of strain L. paracasei LTH 2579 a
b
Monitoring of L. paracasei LTH 2579
Development of a specific detection system for
universal primer 97K permitted the amplification of a 288 bp DNA fragment of the 16S rDNA of the type strain of L. paracasei. The specificity of the species-specific PCR system was confirmed by using the genomic DNA of strains of the species compiled in Table 1. Amplification products were synthesised exclusively with the DNA of the two subspecies of L. paracasei and 14 strains presumptively determined as L. casei as well as 4 nonidentified lactobacilli which therefore were allotted to L. paracasei. The amplification was also operating with crude nucleic acid extracts from colonies taken from agar plates. A species-specific PCR system was also developed to detect L. sakei used as starter for production of the fermented sausages. A target site within the 23S rRNA was chosen which has previously been used to construct the species-specific probe Lbs for L. sakei (HERTEL et al., 1991). The species-specific primer Lsakf was designed (Tab. 2) and combined with group-specific primer 383R resulting in the amplification of a 293 bp DNA fragment of the 23S rDNA. The specificity of the PCR system was confirmed with the genomic DNA of L. sakei LTH 681 and DSM 20017T as well as L. curvatus LTH 684 and DSM 20019 T •
L. paracasei lTH 2579
As it was the aim to follow the fate of L. paracasei LTH 2579, it was intended to identify the minor sequence variations which can be found in closely related strains by applying the subtraction hybridisation technique. Therefore, 9 isolates of L. paracasei were selected as potential subtracter strains. These differed in their physiological properties and shared with the target strain the habitat fruit mash. Upon RAPD-PCR analysis 4 different patterns were identified (Fig. 1) and strains of each group (strains LTH 2570, LTH 2600, LTH 2567, LTH 2552, and LTH 2553) were selected as subtracter strains. Upon subtraction hybridisation with the genomic DNA of the target strain and 5 subtracter strains a non-hybridising 235 bp DNA fragment was obtained. Analysis of the deduced amino acid sequence revealed similarities of 28%, 32%, 52%, and 54% with sequences of 50, 53, 47, and 53 amino acid residues of the phosphopentomutase of Helicobacter pylori, Lactococcus lactis, Bacillus subtilis, and Escherichia coli, respectively. The DNA fragment was used as probe 234L in Southern hybridisations of chromosomal and plasmid DNA of strain
I Q\
l'
N
N
an
M
Fig. 1. RAPD fingerprints of the genomic DNA of L. paracasei strain LTH 2579 (target strain) as well as strains LTH 2570, LTH 2604, LTH 2578, LTH 2600, LTH 2567, LTH 2530, LTH 2553, LTH 2552, and LTH 2568 (potential subtracter strains). The groups I to IV based on the patterns are indicated. HindIII digested A DNA was used as mulecular size marker (M) .
Q
l'
:c ~
an
:c ~
263
III
II 'II:t
Q
\C N
:c ~
00 l'
an
N
:c ~
IV
Q Q
\C N
l' \C
Q ('I')
:c
N
N
~
~
~
an
:c
an
:c
('I')
N
00 \C
N
N
N
~
~
an an
:c ~
an an
:c
an
:c
264
C. BUNTE et al.
LTH 2579 and it was observed that the fragment originates from the chromosome. In addition, with the chromosomal DNA only single hybridisation bands were obtained, thus indicating that the specific DNA fragment is not part of a repetitive element within the chromosome (Fig. 2). For the development of a strain-specific PCR system, primers 2579f and 2579r (Tab. 2) were constructed targeted against the 5'- and 3'-end of the isolated DNA fragment. Amplification with chromosomal DNA of strain LTH 2579 as template resulted in 234 bp PCR product. The specificity of this primer pair was validated using the genomic DNA of the 36 strains of L. paracasei and 51 strains of other Lactobacillus species shown in Table 1. Remarkably, PCR products were obtained also with the genomic DNA of 2 L. paracasei strains (LTH 2066 and LTH 1835) isolated from sauerkraut. Both strains as well as the target strain differed in their plasmid profiles and carbohydrate fermentation patterns. In addition, the RAPD patterns of both strains were rather similar but clearly different from that of the target strain (data not shown). The analysis of these PCR products revealed that the DNA sequences of the three positively reacting strains were identical. Moreover, Southern hybridisation was performed with probe 234L and chromosomal DNA of the subtracter strains as well as
1
2
3
4
L. paracasei LTH 2579, LTH 2066, and DSM 5622 T (data not shown). Hybridisation was observed exclusively with the DNA of strains LTH 2579 and LTH 2066 at a stringency of 70%. Finally, it was shown that the specific PCR system was also applicable to crude nucleic acid extracts from colonies taken from agar plates.
Monitoring of strain LTH 2579 in fermented sausage and feces To follow the fate of strain L. paracasei LTH 2579 in an environment, the species-specific PCR system for L. paracasei was combined with that specific for strain LTH 2579. Fermented sausages were produced using among others L. sakei LTH 681 and L. paracasei LTH 2579 as starter organisms. The lactobacilli were counted at the first day of consumption. Characteristic white and greenish colonies indicated the presence of 1.8 x 107 CFU/g and 1.4 x 108 CFU/g, respectively, of lactobacilli in the sausage. The application of the PCR systems (specific for L. paracasei LTH 2579 and species-specific for L. paracasei and L. sakei) revealed that the white colonies were the strain LTH 2579 and the greenish colonies consisted of the strain L. sakei LTH 681. For monitoring of strain LTH 2579 in feces, three subjects consumed the fermented sausages for 3 days and
5
kbp
23.1
9.4 6.6
4.4
2.3
2.0
Fig. 2. Southern hybridisation of chromosomal DNA of the strain L. paracasei LTH 2579 with the polynucleotide probe 234L. Lane 1 and 5, HindIII-digested A DNA as molecular size marker; lane 2, EcoRI; lane 3, HindIII; lane 4, Pst!'
Monitoring of L. paracasei LTH 2579
265
plasmid profile. Thus, the primers derived from application of the subtraction hybridisation technique are not absolutely specific for one single strain. This observation Sample 1 Sample 2 confirms the findings of WASSILL et al. (1998) that the specificity of strain-specific primers derived by subtracLB LB target tion hybridisation is only warranted for strains included as subtracters or references for specific evaluation. It 1.2 x 10 8 1.8 x 10 8 1.5x10 8 Subject 1 should be noted that an absolute strain specificity can 1.6 x 10 8 1.5 X 10 8 Subject 2 < 3.4 x 10 3 1.3 X 107 1.2 X 107 Subject 3 2.4 x 10 5 never be obtained due to our understanding of the nature of a strain. For example, a point mutation in one gene might strongly affect the physiological properties but can only be genotypically detected if the necessary sequences fecal samples were investigated one day before (sample are known. 1) and at the third day of consumption (sample 2). The One reason for the reduced specificity may be our lack of knowledge about the distribution in the various lactobacilli were counted on LAMVAB agar and the results are compiled in Table 3. For statistical reasons, 13 strains of the conserved sequences within the target DNA different colonies from samples 1, representing all colony fragment. Sequencing of the amplification products did types, were subjected to PCR analysis. L. paracasei was not reveal any differences in the chromosomal DNA identified in two colony types. Upon PCR reaction with fragments of strains L. paracasei LTH 2066, LTH 1835, the primer pair 2579f / 2579r no amplifications were oband LTH 2579. Thus, the specificity of the PCR system tained with the nucleic acid extracts of the 13 colonies. for L. paracasei LTH 2579 could not be further increased Investigation of samples 2 with the PCR-based detection by shifting the primer binding sites. Sequence analysis systems revealed that the target strain LTH 2579 was deof the deduced amino acid sequence revealed that the tectable in the fecal samples of all subjects. The cell 235 bp DNA fragment may be part of a gene coding for counts were determined for lactobacilli and strain LTH the phosphopentomutase. This enzyme is involved in the 2579 (Tab. 3). The confidence intervals for the calculated pyrimidine salvage pathway (NEUHARD and KELLN, LTH 2579 counts were within the range of the counting 1996), and is apparently restricted to only a few strains error common for surface plating and, therefore, are not of L. paracasei. Missing signals upon high stringency (70%) Southern hybridisation with DNA of 6 strains of indicated. Remarkably, the major part of lactobacilli in L. paracasei is consistent with the absence of the gene sethe feces of all subjects consisted of the target strain L. paracasei LTH 2579, the content of non-target lactoquence. The Southern hybridisations showed also that bacilli ranged between 5 and 12 % of total lactobacilli. the DNA fragment is not part of a plasmid or repetitive element in the chromosome, thus supporting its genetic stability and the reliability of the detection system. As Discussion this target sequence might have been acquired by horizontal gene transfer, the combination of the PCR system Lactobacilli such as Lactobacillus acidophilus, L. johnwith a species-specific system was used to exclude false sonii, L. rhamnosus, and L. paracasei are used in starter positive results. cultures or as probiotics, are consumed with the food In this study it was demonstrated that the specific and may interact with the intestinal flora or even to the PCR-based detection system with subtraction hybridisahuman host. In previous studies (HAMMES et al., 1997) it tion derived-primers is applicable to monitor the fate of had been shown that bacteria involved in food fermentaspecific bacteria in the food chain. For this purpose, the tion share some essential properties with pro biotic bactefood associated organism L. paracasei LTH 2579 was ria, e.g. tolerance to hydrochloric acid (pH 1.5-2.5) and chosen as model. This strain is highly resistant to low pH bile, and adhesion to enterocytic cells. For technological. and bile at physiological relevant concentrations. In the reasons as well as for evaluating the effect in the intespast several attempts were made to detect a specific strain in fecal samples by characterisation of the re-isotine, knowledge is required as to the fate of these organlated strain. For example, JACOBSEN et al. (1999) have isms after inoculation in the food till passage through the used a combined detection method based on phenotypic intestines. For this purpose, a strain-specific detection and two genotypic characteristics to investigate the surmethod is needed, as the implanted lactobacilli share commonly the habitat with very similar species and vival of various strains of lactobacilli including dairy strains. We consider our approach less time-consuming, strains. The developed PCR systems proved to fulfil this requirement and were shown to permit the specific detecas it works also with crude nucleic acid extracts from tion of a specific strain as well as specific species. Validacolonies and enables to reliably detect the target strain tion of the specificity for L. paracasei LTH 2579 revealed within 6 hours. The quantification of L. paracasei LTH 2579 in the feces revealed that this food fermenting that PCR products are obtained also with the DNA of two isolates of sauerkraut. It can be excluded that these strain survives the passage through the intestinal tract at high numbers. We are aware that the data are based on strains are identical with our target strain, as they fell the results obtained with three volunteers only and thus into a different RAPD-PCR group and each exhibited an the conclusion cannot be generalised. Nevertheless, it unique carbohydrate fermentation pattern as well as
Table 3. Numbers (CFU/g) of lactobacilli (LB) and of strain L. paracasei LTH 2579 (target) in fecal samples.
266
e. BUNTE et a!.
was demonstrated that bacteria involved in food fermentations can reach the large bowel alive and thus may interact with the resident microflora as discussed for probiotic strains. Acknowledgements We thank e. Fedorow for excellent technical assistance. We are indebted to U. Gehring and H. J. Buckenhiiskes (Gewiirzmiiller) for providing fermented sausages and to L. Wassill and W. Ludwig (Technische Universitat Miinchen) for helpful discussion. This work was supported by the "Ministerium fiir Wissenschaft, Forschung und Kunst Baden-Wiirttemberg" and Gewiirzmiiller GmbH, Stuttgart, Germany.
References BjOURSEN, A. J., COOPER, J. E.: Isolation of Rhizobium loti strain-specific DNA sequences by subtraction hybridization. App!. Environ. Microbio!. 54, 2852-2855 (1988). CAVADINI, e., HERTEL, e., HAMMES, W. P.: Stable expression of the lysostaphin gene in meat lactobacilli by introducing deletions within the prosequence. System. App!. Microbio!. 19, 21-27 (1996). CROWTHER, J. S.: Transport and storage of faeces for bacteriological examination. J. App!. Bacterio!' 34, 477-483 (1971). DE MAN, J. e., ROGOSA, M., SHARPE, M. E.: A medium for the cultivation of lacto bacilli. J. App!. Bacterio!' 23, 130-135 (1960). HAMMES, W. P., HALLER, D., BRASSART, D., BODE, e.: Traditional starter cultures as probiotics. Microeco!' Ther. 26, 97-114 (1997). HAMMES, W. P., BANTLE ON, A., MIN S.: Lactic acid bacteria in meat fermentation. FEMS Microbio!. Rev. 87, 165-174 (1990). HARTEMINK, R., DOMENECH, V. R., ROMBOUTS, EM.: LAMVAB - A new selective medium for the isolation of lactobacilli from faeces. J. Microbio!. Methods 29, 77-84 (1997). HENRIKSEN, S. D.: Serotyping of bacteria. Methods Microbiol. 12,1-13 (1978). HERTEL, C., LUDWIG, W., OBST, M., VOGEL, R., HAMMES, W. P., SCHLEIFER K. H.: 23S rRNA-targeted oligonucleotide probes for the rapid identification of meat lactobacilli. System. App!. Microbio!. 14, 173-177 (1991). JACOBSEN, e. N., ROSENFELDT NIELSEN, V., HAYFORD, A. E., M0LLER, P. L., MICHAELSEN, K. E, PJERREGAARD, A., SANDSTROM, B., TVEDE, M., JAKOBSEN, M.: Screening of probiotic activities of forty-seven strains of Lactobacillus spp. by in
vitro techniques and evaluation of the colonization ability of five selected strains in humans. App!. Environ. Microbio!. 65,4949-4956 (1999). MATHESON, V. G., MUNAKATA-MARR, J., HOPKINS, G. D., McCARTY, P. L., TIEDjE J. M., FORNEY, L. J.: A novel means to develop strain-specific DNA probes for detecting bacteria in the environment. App!. Environ. Microbia!. 63, 2863-2869 (1997). NEUHARD, J., KELLN, R. A.: Biosynthesis and conversions of pyrimidines, pp. 580-599. In: Escherichia coli and Salmonella (Ee. NEIDHARDT ed.) 2nd ed., Washington, ASM Press 1996. SAHAI, H., KURSHID, A.: Statistics in epidemiology. Boca Raton, CRC press 1996. SCHMIDT, G., HERTEL, e., H AMMES W. P.: Molecular characterization of the dnaK operon of Lactobacillus sakei LTH681. System. App!. Microbio!. 22, 321-328 (1999). SPRINGER, N., LUDWIG, W., DROZANSKI, V., AMANN, R ., SCHLEIFER, K.-H.: The phylogenetic status of Sarcobium lyticum, an obligate intracellular bacterial parasite of small amoebae. FEMS Microbio!. Lett. 96, 199-202 (1992). TILSALA-TIMISjARVI, A., ALATOSSAVA, T.: Strain-specific identification of probiotic Lactobacillus rhamnosus with randomly amplified polymorphic DNA-derived PCR primers. App!. Environ. Microbio!. 64,4816-4819 (1998). TORRIANI, S., VAN REENEN, e. A., KLEIN, G., REUTER, G., DELLAGLIO, E and DICKS, L. M. T.: Lactobacillus curvatus subsp. curvatus subsp. nov. and Lactobacillus curvatus subsp. melibiosus subsp. nov. and Lactobacillus sake subsp. sake subsp. nov. and Lactobacillus sake subsp. carnosus subsp. nov., new subspecies of Lactobacillus curvatus Abo-Elnaga and Kandler 1965 and Lactobacillus sake Katagiri, Kitahara, and Fukami 1934 (Klein et a!. 1996, Emended Descriptions), respectively. Int. J. Syst. Bacteria!' 46, 1158-1163 (1996). VANDAMME, P., POT, B., GILLIS, M., DE VOS, P., KERSTERS, K., SWINGS, J.: Poly phasic taxonomy, a consensus approach to bacterial systematics. Microbia!. Rev. 60,407-438 (1996). WASSILL, L., LUDWIG, W., SCHLEIFER, K. H.: Development of a modified subtraction hybridization technique and its application for the design of strain specific PCR systems for lactococci. FEMS Microbio!. Lett. 95, 87-94 (1998).
Corresponding author: CHRISTIAN HERTEL, Institut fiir Lebensmitteltechnologie, Universitat Hohenheim, Garbenstr. 28, D - 70599 Stuttgart, Germany Phone: +49 711 459 4255; Fax: +49 711 4594199 e-mail: [email protected]
© Urban & Fischer Verlag
Monitoring and Survival of Lactobacillus paracasei LTH 2579 in Food and the Human Intestinal Tract CHRISTOPH BUNTE, CHRISTIAN HERTEL, and WALTER P. HAMMES
Institut fur Lebensmitteltechnologie, Universitat Hohenheim, Stuttgart, Germany Received April 3, 2000
Summary A PCR-based detection system specific for Lactobacillus paracasei LTH 25 79 was developed and applied to follow the fate of the strain in complex ecosystems. This strain was isolated from fruit mash and was characterised as being highly resistant to low pH and bile at concentrations as they occur in the human digestive tract. The application of the subtraction hybridisation technique permitted to identify a 235 bp chromosomal DNA fragment of strain LTH 2579. Based on this target sequence a specific PCR system was developed and combined with the species-specific PCR system for L. paracasei. This combination of PCR based detection systems was successfully applied to monitor L. paracasei LTH 2579 in fermented sausages which were inoculated with this strain (2.0 x 10 7 CFU/g) together with the strongly competitive L. sakei LTH 681 (1.0 x 10 6 CFU/g). At the time of consumption of the sausages the respective counts were 1.8 x 10 7 and 1.4 x 10 8 CFU/g. After consumption of the sausages by three volunteers L. paracasei LTH 2579 was recovered from fecal samples. The counts determined for the strain ranged between 1.2 x 10 7 and 1.5 x 10 8 CFU/g of feces. The fortuitous lactobacilli constituted a share of 5-12% of the lactobacilli in the fecal flora. Key words: strain-specific PCR - species-specific PCR - subtraction hybridisation - Lactobacillus para-
casei - Lactobacillus sakei - intestinal flora - fermented food
Introduction To monitor the fate of a bacterial strain implanted into complex ecosystems, unique phenotypic or genotypic markers are needed for its reliable and specific detection. The phenotypic approach, e.g. the application of serological techniques (HENRIKSEN, 1978), depends on the expression of the phenotypic marker and may thus be affected by environmental factors. On the other hand, genotypic detection methods with taxonomic resolution to the strain level (VANDAMME et al., 1996) have been proven more useful, e.g. restriction fragment length polymorphism (RFLP) and random amplified polymorphic DNA (RAPD-PCR). A further approach is the detection of strain-specific DNA sequences by applying hybridisation or PCR techniques with specific probes or primers. These unique sequences can be derived, for example, from RAPD fragments, as it was recently described for the detection of Lactobacillus rhamnosus Lc 1/3 (TILSALA-TIMISJARVI and ALATOSSAVA, 1998) or may be obtained by the subtraction hybridisation technique (BJOURSEN and COOPER, 1988; MATHESON et al., 1997). The applicability of this method to develop a strain-spe0723-2020/00/23/02-260 $ 12.00/0
cific PCR-based detection system has recently been demonstrated by WASSILL et al. (1998). It was the purpose of our investigations to make use of this technique and to develop a specific PCR system for Lactobacillus paracasei LTH 2579. This strain was isolated from fermenting fruit mash and its properties were compared with those of intestinal lactobacilli (HAMMES et al., 1997). It had been shown that strain LTH 2579 is resistant to low pH and bile at concentrations as they occur in the human digestive tract. The strain also adhered to Caco 2-cell cultures. It was our aim to establish a model for monitoring bacterial strains such as starter or probiotic organisms in both fermented food and fecal samples.
Materials and Methods Bacterial strains. growth conditions. and physiological characterisation
The strains used in this study were Lactobacillus sakei DSM 20017T , Lactobacillus curvatus DSM 20019 T , L. sakei LTH 681
Monitoring of L. paracasei LTH 2579 and L. curvatus LTH 684 formerly referred to as strains Ls12 and Lc3 (HAMMES et al., 1990), respectively, and those strains compiled in Table 1. Lactobacilli were grown in MRS medium (DE MAN et al., 1960) at 30°C. For selective detection of intestinallactobacilli, LAMVAB medium (HARTEMINK et al., 1997) was used, and incubations were performed at 37°C under anaerobic conditions (Anaerocult A, Merck). Physiological characterisation of the lactobacilli was performed using the API 50 CH test system (BioMerieux) together with the API CHL medium.
Production and microbial analysis of fermented sausages For the production of 1 kg sausage mixture, 0.90 kg frozen pork meat and 0.10 kg frozen pork belly (40 % lean) were cut and mixed at ambient temperature. Glucose (2 g), nitrite curing salt (26 g), and spice preparation (8 g) were added. Lyophilised starter organisms were suspended in 50 ml water and added to the meat mixture to obtain final cell counts (CFU/g) of 1.0 x 10 6 of L. sakei LTH 681, 8.0 x 10 6 of Staphylococcus carnosus LTH 3127, 1.5 x 10 6 of Kocuria varians LTH 2864, and 2.0 x 10 7 of L. paracasei LTH 2579. Sausages of ca. 68 g were filled in cellulose casings of 20 mm diameter and incubated in ripening chambers (Siidtronic) for 122.5 h. The following ripening conditions were applied: 24 h at 24°C (90-92 % r.H.), 2 h at 24-28 °C (50% r.H.), 30 min at 24-28 °C (smoking), and 96 h at 14-15 °C (75-80% r.H.) . The loss of weight during ripening was 25%. For microbial analysis samples of 10 g of sausage were homogenised with the aid of a stomacher in 90 ml of dilution medium containing per I: 8.5 g NaCI, 1.0 g peptone. Microbial counts of LAB were determined by surface plating on MRS medium supplemented with 0.025 gIl bromocresol green. Microbial analysis of fecal samples Fecal samples were taken from 3 male subjects at 1 day before (sample 1) and on the third day (sample 2) of consumption of 100 glday (subjects 2 and 3) or 200 glday (subject 1). Samples of ca. 0.2 g were transferred to pre-weighed tubes containing 2 ml of a pre-reduced cryoprotective broth (CROWTHER, 1971) which was supplemented with 0.25 gIl cysteine-HCI and 0.15 mgll resazurine. The samples were homogenised, and immediately frozen and stored below -20°C. Cell counts were determined within 3 weeks by decimal dilute in a pre-reduced dilution medium (8.5 gil NaCl, 1.0 gil peptone, and 0.1 gil cysteine) and plating on LAMVAB medium. All anaerobic manipulations were carried out in an anaerobic chamber under an atmosphere of 85% N z, 10% CO 2 , 5% H z. Plates were incubated for 3 days at 37°C under anaerobic conditions. Viable cell counts were expressed as CFU per gram wet weight feces. For counting of L. paracasei LTH 2579, one colony of each colony type was investigated by application of the strain-specific as well as the L. paracasei-specific PCR detection system. The colonies of the type yielding a double positive result were counted, a number of 20 was randomly picked and also subjected to PCR analysis to confirm the homogeneity of the colony type. The cell count of L. paracasei LTH 2579 was finally determined applying the hypergeometric distribution with a level of significance a =5% (SAHAI and KURSHID, 1996). DNA isolation and preparation of crude nucleic acid extracts For isolation of the genomic DNA used as template for PCR, the High Pure PCR Template Preparation Kit (Roche Diagnostics GmbH) was applied. Chromosomal and plasmid DNA of lactobacilli were isolated as described previously (SCHMIDT et al., 1999; CAVADINI et al., 1996). Crude nucleic acid extracts were prepared by resuspending a colony in 40 fli of bidistilled water in a reaction tube. The suspension was heated for 10 min at 98°C and centrifuged at 18,OOOx g for 2 min. An aliquot of the supernatant was subjected to PCR.
261
Design of specific primers To construct the primer specific for Lactobacillus paracasei, 16S rRNA sequences of Lactobacillus paracasei subsp. paracasei, L. paracasei subsp. tolerans, L. casei, L. rhamnosus, and L. zeae were aligned and compared to obtain specific target sites. The EMBL accession numbers of the sequences are D79212, D16550, D16551, D16552, and D86516, respectively. For verification of the specificity, the Probe Match tool of the ARB program package (SPRINGER et al., 1992) was used together with a database containing all deposited complete primary structures of 16S rRNA. For construction of the primer specific for L. sakei, the site within the 23S rDNA which was used previously to design the probe Lbs (HERTEL et al., 1991) was chosen as target site for the primer. To increase the number of mismatches and to place two of three mismatches at the 3' region of the primer, the target site of the primer was shifted 13 bases to the 5' end of the 23S rRNA. For construction of specific primers for L. paracasei LTH 2579 the subtraction hybridisation technique was used as described by WASSIll et al. (1998). The genomic DNA of strain LTH 2579 served as target DNA and that of the L. paracasei strains LTH 2570, LTH 2552, LTH 2553, LTH 2567, and LTH 2600 as subtracter DNA. One DNA fragment of 235 bp was isolated and sequenced applying the dideoxynucleotide chain termination method and universal primers. Based on the sequence specific primers were designed for detection of the strain L. paracasei LTH 2579. Specific and universal polymerase chain reactions All PCR amplifications were carried out on a GeneAmp 2400 Thermocycler (Perkin Elmer) in total volumes of 25 fll. The reaction mixture contained 0.2 mM of each deoxynucleotide triphosphate, 0.5 flM of each primer, PCR buffer (1.5 mM MgCI2' 50 mM KCI, and 10 mM Tris-HCI, pH 9.0), 1.5 U Taq polymerase (Amersham Pharmacia Biotech), and 7 fli of the nucleic acid extract or 50 ng genomic DNA. The following amplification profiles were used: i) for specific detection of L. paracasei: 1 cycle of 2 min at 94°C; 32 cycles of 1 min at 94 °C, 1 min at 49°C, and 30 sec at 72 °C; 1 cycle of 5 min at 72 °C; ii) for specific detection of L. sakei: 1 cycle of 2 min at 94°C; 35 cycles of 1 min at 94 °C, 1 min at 57°C, and 20 sec at 72 °C; 1 cycle of 5 min at 72 °C; iii) for the control reaction with universal primers 616VII and lOOK : 1 cycle of 2 min at 94°C; 32 cycles of 30 sec at 94 °C, 45 sec at 51°C, and 90 sec at 72 °C; 1 cycle of 5 min at 72 0C. For specific detection of the strain L. paracasei LTH 2579 the reaction mix was modified as follows: 0.1 mM of each deoxynucleotide triphosphate, 2.5 U Taq polymerase (Amersham Pharmacia Biotech), and 12 fli of the nucleic acid extract. The amplification profile was the following: 1 cycle of 2 min at 94°C; 35 cycles of 1 min at 94 °C, 1 min at 68°C, and 30 sec at 72 °C; 1 cycle of 5 min at 72 0C. The total contents of the PCR reaction vials were subjected to agarose gel electrophoresis and visualised by staining with ethidium bromide. RAPD polymerase chain reaction RAPD-PCR was performed using primer OPL-02 (TORRIANI et al., 1996). The reaction mix was modified by using 0.2 flM of each primer and 2.5 U Taq polymerase. The amplification profile was the following: 1 cycle of 2 min at 94°C, 45 cycles of 1 min at 94 °C, 1 min at 34°C and 1 min at 72 °C, 1 cycle of 5 min at 72 0C. 10 fli of the content of each PCR reaction vial was subjected to agarose gel electrophoresis and visualised by staining with ethidium bromide. Southern hybridisation Genomic DNA (1 fig) of strain LTH 2579 was digested with the restriction enzymes EcoRI, HindIII, or Pst!. Plasmid DNA
262
C. BUNTE et al.
and the genomic restriction fragments were separated by agarose gel electrophoresis and transferred to a nylon membrane (Quiabrane, Qiagen) using a vacuum blot apparatus (Amersham Pharmacia Biotech) following the protocol of the supplier. After blotting the DNA was immobilised by incubating at 80°C for 30 min. For hybridisation the digoxygenin-labelled probe 234L was constructed using a PCR DIG Probe Synthesis Kit (Roche Diagnostics GmbH), primer pair 2579f / 2579r and genomic DNA of strain L. paracasei LTH 2579 following the instructions of the supplier. Hybridisation (over night) and stringent washing were carried out at 64°C. Detection was performed using the Colorimetric DIG detection kit (Roche Diagnostics GmbH) according to the instruction of the supplier.
Computer analysis The DNA sequences were analysed with the aid of the software DNASIS v2.5 (Hitachi Software Engineering Co.). Amino acid sequences were compared to the Swissprot database (Swiss
Institute of Bioinformatics, Geneva) and to the Protein information resource (National Biomedical Research Foundation, Washington) using BLAST.
Results Development of species-specific detection systems To develop a PCR-based system for specific detection of L. paracasei, the 16S rDNA was chosen as target molecule. Comparative analysis of the 16S rRNA sequences revealed a region that was used as target site for the species-specific primer Lpacaf (Tab. 2). The Computer-based validation of the specificity showed that all potential primer target hybrids contained at least one weighted mismatch. Combination of this primer with the
Table 1. Specificity of the PCR with primer pairs Lpacaf /97 K and 2579f / 2579r. Strain
Species
PCR with primer pair
Source
Lpacaf /97 K
L. L. L. L. L.
casei zeae rhamnosus paracasei subsp. tolerans paracasei subsp. paracasei
L. casei
Lactobacillus spec. a
DSM 20011 T DSM 20178 T DSM 20021 T DSM 20258 T DSM 5622 T DSM20008 DSM20020 DSM20207 DSM 20312 LTH2579 (2 strains) (2 strains) (7 strains) (1 strain) (8 strains) (1 strain) (5 strains) (4 strains) (48 strains)
2579f / 2579r
cheese corn steep liquor milk
+ + + + + + + + + + + + + + +
milk saliva milk feces apple mash sauerkraut cherry mash pear mash plum mash sauerkraut beer unknown feces a feces a
+ +
isolated on LAMVAB agar from 22 subjects
Table 2. Specific and universal primers used in the PCR systems. Primer
Sequence (5'-3')
Specificity
Position
Lpacaf 97K Lsakf 383R 2579f 2579r 616VII lOOK
CCGAGATTCAACATGG CTGCTGCCTCCCGTA CCCAATACTTTTAATCGA GTTCCGCCGTACTCA GATTTAGTCTGCACACGC CCTTCACTGTCATCATCC AGAGTTTGATYMTGGCTCAG GGTTACCTTGTTACGACTT
L. paracasei universal L. sakei group-specific L. paracasei LTH 2579 L. paracasei LTH 2579 universal universal
88 a,b 357 a,b 128 a,' 383 a,c 2d 235 d 7 a,b 1491 a,b
according to E.coli numbering 16S rDNA c 23S rDNA d 235 bp DNA fragment of strain L. paracasei LTH 2579 a
b
Monitoring of L. paracasei LTH 2579
Development of a specific detection system for
universal primer 97K permitted the amplification of a 288 bp DNA fragment of the 16S rDNA of the type strain of L. paracasei. The specificity of the species-specific PCR system was confirmed by using the genomic DNA of strains of the species compiled in Table 1. Amplification products were synthesised exclusively with the DNA of the two subspecies of L. paracasei and 14 strains presumptively determined as L. casei as well as 4 nonidentified lactobacilli which therefore were allotted to L. paracasei. The amplification was also operating with crude nucleic acid extracts from colonies taken from agar plates. A species-specific PCR system was also developed to detect L. sakei used as starter for production of the fermented sausages. A target site within the 23S rRNA was chosen which has previously been used to construct the species-specific probe Lbs for L. sakei (HERTEL et al., 1991). The species-specific primer Lsakf was designed (Tab. 2) and combined with group-specific primer 383R resulting in the amplification of a 293 bp DNA fragment of the 23S rDNA. The specificity of the PCR system was confirmed with the genomic DNA of L. sakei LTH 681 and DSM 20017T as well as L. curvatus LTH 684 and DSM 20019 T •
L. paracasei lTH 2579
As it was the aim to follow the fate of L. paracasei LTH 2579, it was intended to identify the minor sequence variations which can be found in closely related strains by applying the subtraction hybridisation technique. Therefore, 9 isolates of L. paracasei were selected as potential subtracter strains. These differed in their physiological properties and shared with the target strain the habitat fruit mash. Upon RAPD-PCR analysis 4 different patterns were identified (Fig. 1) and strains of each group (strains LTH 2570, LTH 2600, LTH 2567, LTH 2552, and LTH 2553) were selected as subtracter strains. Upon subtraction hybridisation with the genomic DNA of the target strain and 5 subtracter strains a non-hybridising 235 bp DNA fragment was obtained. Analysis of the deduced amino acid sequence revealed similarities of 28%, 32%, 52%, and 54% with sequences of 50, 53, 47, and 53 amino acid residues of the phosphopentomutase of Helicobacter pylori, Lactococcus lactis, Bacillus subtilis, and Escherichia coli, respectively. The DNA fragment was used as probe 234L in Southern hybridisations of chromosomal and plasmid DNA of strain
I Q\
l'
N
N
an
M
Fig. 1. RAPD fingerprints of the genomic DNA of L. paracasei strain LTH 2579 (target strain) as well as strains LTH 2570, LTH 2604, LTH 2578, LTH 2600, LTH 2567, LTH 2530, LTH 2553, LTH 2552, and LTH 2568 (potential subtracter strains). The groups I to IV based on the patterns are indicated. HindIII digested A DNA was used as mulecular size marker (M) .
Q
l'
:c ~
an
:c ~
263
III
II 'II:t
Q
\C N
:c ~
00 l'
an
N
:c ~
IV
Q Q
\C N
l' \C
Q ('I')
:c
N
N
~
~
~
an
:c
an
:c
('I')
N
00 \C
N
N
N
~
~
an an
:c ~
an an
:c
an
:c
264
C. BUNTE et al.
LTH 2579 and it was observed that the fragment originates from the chromosome. In addition, with the chromosomal DNA only single hybridisation bands were obtained, thus indicating that the specific DNA fragment is not part of a repetitive element within the chromosome (Fig. 2). For the development of a strain-specific PCR system, primers 2579f and 2579r (Tab. 2) were constructed targeted against the 5'- and 3'-end of the isolated DNA fragment. Amplification with chromosomal DNA of strain LTH 2579 as template resulted in 234 bp PCR product. The specificity of this primer pair was validated using the genomic DNA of the 36 strains of L. paracasei and 51 strains of other Lactobacillus species shown in Table 1. Remarkably, PCR products were obtained also with the genomic DNA of 2 L. paracasei strains (LTH 2066 and LTH 1835) isolated from sauerkraut. Both strains as well as the target strain differed in their plasmid profiles and carbohydrate fermentation patterns. In addition, the RAPD patterns of both strains were rather similar but clearly different from that of the target strain (data not shown). The analysis of these PCR products revealed that the DNA sequences of the three positively reacting strains were identical. Moreover, Southern hybridisation was performed with probe 234L and chromosomal DNA of the subtracter strains as well as
1
2
3
4
L. paracasei LTH 2579, LTH 2066, and DSM 5622 T (data not shown). Hybridisation was observed exclusively with the DNA of strains LTH 2579 and LTH 2066 at a stringency of 70%. Finally, it was shown that the specific PCR system was also applicable to crude nucleic acid extracts from colonies taken from agar plates.
Monitoring of strain LTH 2579 in fermented sausage and feces To follow the fate of strain L. paracasei LTH 2579 in an environment, the species-specific PCR system for L. paracasei was combined with that specific for strain LTH 2579. Fermented sausages were produced using among others L. sakei LTH 681 and L. paracasei LTH 2579 as starter organisms. The lactobacilli were counted at the first day of consumption. Characteristic white and greenish colonies indicated the presence of 1.8 x 107 CFU/g and 1.4 x 108 CFU/g, respectively, of lactobacilli in the sausage. The application of the PCR systems (specific for L. paracasei LTH 2579 and species-specific for L. paracasei and L. sakei) revealed that the white colonies were the strain LTH 2579 and the greenish colonies consisted of the strain L. sakei LTH 681. For monitoring of strain LTH 2579 in feces, three subjects consumed the fermented sausages for 3 days and
5
kbp
23.1
9.4 6.6
4.4
2.3
2.0
Fig. 2. Southern hybridisation of chromosomal DNA of the strain L. paracasei LTH 2579 with the polynucleotide probe 234L. Lane 1 and 5, HindIII-digested A DNA as molecular size marker; lane 2, EcoRI; lane 3, HindIII; lane 4, Pst!'
Monitoring of L. paracasei LTH 2579
265
plasmid profile. Thus, the primers derived from application of the subtraction hybridisation technique are not absolutely specific for one single strain. This observation Sample 1 Sample 2 confirms the findings of WASSILL et al. (1998) that the specificity of strain-specific primers derived by subtracLB LB target tion hybridisation is only warranted for strains included as subtracters or references for specific evaluation. It 1.2 x 10 8 1.8 x 10 8 1.5x10 8 Subject 1 should be noted that an absolute strain specificity can 1.6 x 10 8 1.5 X 10 8 Subject 2 < 3.4 x 10 3 1.3 X 107 1.2 X 107 Subject 3 2.4 x 10 5 never be obtained due to our understanding of the nature of a strain. For example, a point mutation in one gene might strongly affect the physiological properties but can only be genotypically detected if the necessary sequences fecal samples were investigated one day before (sample are known. 1) and at the third day of consumption (sample 2). The One reason for the reduced specificity may be our lack of knowledge about the distribution in the various lactobacilli were counted on LAMVAB agar and the results are compiled in Table 3. For statistical reasons, 13 strains of the conserved sequences within the target DNA different colonies from samples 1, representing all colony fragment. Sequencing of the amplification products did types, were subjected to PCR analysis. L. paracasei was not reveal any differences in the chromosomal DNA identified in two colony types. Upon PCR reaction with fragments of strains L. paracasei LTH 2066, LTH 1835, the primer pair 2579f / 2579r no amplifications were oband LTH 2579. Thus, the specificity of the PCR system tained with the nucleic acid extracts of the 13 colonies. for L. paracasei LTH 2579 could not be further increased Investigation of samples 2 with the PCR-based detection by shifting the primer binding sites. Sequence analysis systems revealed that the target strain LTH 2579 was deof the deduced amino acid sequence revealed that the tectable in the fecal samples of all subjects. The cell 235 bp DNA fragment may be part of a gene coding for counts were determined for lactobacilli and strain LTH the phosphopentomutase. This enzyme is involved in the 2579 (Tab. 3). The confidence intervals for the calculated pyrimidine salvage pathway (NEUHARD and KELLN, LTH 2579 counts were within the range of the counting 1996), and is apparently restricted to only a few strains error common for surface plating and, therefore, are not of L. paracasei. Missing signals upon high stringency (70%) Southern hybridisation with DNA of 6 strains of indicated. Remarkably, the major part of lactobacilli in L. paracasei is consistent with the absence of the gene sethe feces of all subjects consisted of the target strain L. paracasei LTH 2579, the content of non-target lactoquence. The Southern hybridisations showed also that bacilli ranged between 5 and 12 % of total lactobacilli. the DNA fragment is not part of a plasmid or repetitive element in the chromosome, thus supporting its genetic stability and the reliability of the detection system. As Discussion this target sequence might have been acquired by horizontal gene transfer, the combination of the PCR system Lactobacilli such as Lactobacillus acidophilus, L. johnwith a species-specific system was used to exclude false sonii, L. rhamnosus, and L. paracasei are used in starter positive results. cultures or as probiotics, are consumed with the food In this study it was demonstrated that the specific and may interact with the intestinal flora or even to the PCR-based detection system with subtraction hybridisahuman host. In previous studies (HAMMES et al., 1997) it tion derived-primers is applicable to monitor the fate of had been shown that bacteria involved in food fermentaspecific bacteria in the food chain. For this purpose, the tion share some essential properties with pro biotic bactefood associated organism L. paracasei LTH 2579 was ria, e.g. tolerance to hydrochloric acid (pH 1.5-2.5) and chosen as model. This strain is highly resistant to low pH bile, and adhesion to enterocytic cells. For technological. and bile at physiological relevant concentrations. In the reasons as well as for evaluating the effect in the intespast several attempts were made to detect a specific strain in fecal samples by characterisation of the re-isotine, knowledge is required as to the fate of these organlated strain. For example, JACOBSEN et al. (1999) have isms after inoculation in the food till passage through the used a combined detection method based on phenotypic intestines. For this purpose, a strain-specific detection and two genotypic characteristics to investigate the surmethod is needed, as the implanted lactobacilli share commonly the habitat with very similar species and vival of various strains of lactobacilli including dairy strains. We consider our approach less time-consuming, strains. The developed PCR systems proved to fulfil this requirement and were shown to permit the specific detecas it works also with crude nucleic acid extracts from tion of a specific strain as well as specific species. Validacolonies and enables to reliably detect the target strain tion of the specificity for L. paracasei LTH 2579 revealed within 6 hours. The quantification of L. paracasei LTH 2579 in the feces revealed that this food fermenting that PCR products are obtained also with the DNA of two isolates of sauerkraut. It can be excluded that these strain survives the passage through the intestinal tract at high numbers. We are aware that the data are based on strains are identical with our target strain, as they fell the results obtained with three volunteers only and thus into a different RAPD-PCR group and each exhibited an the conclusion cannot be generalised. Nevertheless, it unique carbohydrate fermentation pattern as well as
Table 3. Numbers (CFU/g) of lactobacilli (LB) and of strain L. paracasei LTH 2579 (target) in fecal samples.
266
e. BUNTE et a!.
was demonstrated that bacteria involved in food fermentations can reach the large bowel alive and thus may interact with the resident microflora as discussed for probiotic strains. Acknowledgements We thank e. Fedorow for excellent technical assistance. We are indebted to U. Gehring and H. J. Buckenhiiskes (Gewiirzmiiller) for providing fermented sausages and to L. Wassill and W. Ludwig (Technische Universitat Miinchen) for helpful discussion. This work was supported by the "Ministerium fiir Wissenschaft, Forschung und Kunst Baden-Wiirttemberg" and Gewiirzmiiller GmbH, Stuttgart, Germany.
References BjOURSEN, A. J., COOPER, J. E.: Isolation of Rhizobium loti strain-specific DNA sequences by subtraction hybridization. App!. Environ. Microbio!. 54, 2852-2855 (1988). CAVADINI, e., HERTEL, e., HAMMES, W. P.: Stable expression of the lysostaphin gene in meat lactobacilli by introducing deletions within the prosequence. System. App!. Microbio!. 19, 21-27 (1996). CROWTHER, J. S.: Transport and storage of faeces for bacteriological examination. J. App!. Bacterio!' 34, 477-483 (1971). DE MAN, J. e., ROGOSA, M., SHARPE, M. E.: A medium for the cultivation of lacto bacilli. J. App!. Bacterio!' 23, 130-135 (1960). HAMMES, W. P., HALLER, D., BRASSART, D., BODE, e.: Traditional starter cultures as probiotics. Microeco!' Ther. 26, 97-114 (1997). HAMMES, W. P., BANTLE ON, A., MIN S.: Lactic acid bacteria in meat fermentation. FEMS Microbio!. Rev. 87, 165-174 (1990). HARTEMINK, R., DOMENECH, V. R., ROMBOUTS, EM.: LAMVAB - A new selective medium for the isolation of lactobacilli from faeces. J. Microbio!. Methods 29, 77-84 (1997). HENRIKSEN, S. D.: Serotyping of bacteria. Methods Microbiol. 12,1-13 (1978). HERTEL, C., LUDWIG, W., OBST, M., VOGEL, R., HAMMES, W. P., SCHLEIFER K. H.: 23S rRNA-targeted oligonucleotide probes for the rapid identification of meat lactobacilli. System. App!. Microbio!. 14, 173-177 (1991). JACOBSEN, e. N., ROSENFELDT NIELSEN, V., HAYFORD, A. E., M0LLER, P. L., MICHAELSEN, K. E, PJERREGAARD, A., SANDSTROM, B., TVEDE, M., JAKOBSEN, M.: Screening of probiotic activities of forty-seven strains of Lactobacillus spp. by in
vitro techniques and evaluation of the colonization ability of five selected strains in humans. App!. Environ. Microbio!. 65,4949-4956 (1999). MATHESON, V. G., MUNAKATA-MARR, J., HOPKINS, G. D., McCARTY, P. L., TIEDjE J. M., FORNEY, L. J.: A novel means to develop strain-specific DNA probes for detecting bacteria in the environment. App!. Environ. Microbia!. 63, 2863-2869 (1997). NEUHARD, J., KELLN, R. A.: Biosynthesis and conversions of pyrimidines, pp. 580-599. In: Escherichia coli and Salmonella (Ee. NEIDHARDT ed.) 2nd ed., Washington, ASM Press 1996. SAHAI, H., KURSHID, A.: Statistics in epidemiology. Boca Raton, CRC press 1996. SCHMIDT, G., HERTEL, e., H AMMES W. P.: Molecular characterization of the dnaK operon of Lactobacillus sakei LTH681. System. App!. Microbio!. 22, 321-328 (1999). SPRINGER, N., LUDWIG, W., DROZANSKI, V., AMANN, R ., SCHLEIFER, K.-H.: The phylogenetic status of Sarcobium lyticum, an obligate intracellular bacterial parasite of small amoebae. FEMS Microbio!. Lett. 96, 199-202 (1992). TILSALA-TIMISjARVI, A., ALATOSSAVA, T.: Strain-specific identification of probiotic Lactobacillus rhamnosus with randomly amplified polymorphic DNA-derived PCR primers. App!. Environ. Microbio!. 64,4816-4819 (1998). TORRIANI, S., VAN REENEN, e. A., KLEIN, G., REUTER, G., DELLAGLIO, E and DICKS, L. M. T.: Lactobacillus curvatus subsp. curvatus subsp. nov. and Lactobacillus curvatus subsp. melibiosus subsp. nov. and Lactobacillus sake subsp. sake subsp. nov. and Lactobacillus sake subsp. carnosus subsp. nov., new subspecies of Lactobacillus curvatus Abo-Elnaga and Kandler 1965 and Lactobacillus sake Katagiri, Kitahara, and Fukami 1934 (Klein et a!. 1996, Emended Descriptions), respectively. Int. J. Syst. Bacteria!' 46, 1158-1163 (1996). VANDAMME, P., POT, B., GILLIS, M., DE VOS, P., KERSTERS, K., SWINGS, J.: Poly phasic taxonomy, a consensus approach to bacterial systematics. Microbia!. Rev. 60,407-438 (1996). WASSILL, L., LUDWIG, W., SCHLEIFER, K. H.: Development of a modified subtraction hybridization technique and its application for the design of strain specific PCR systems for lactococci. FEMS Microbio!. Lett. 95, 87-94 (1998).
Corresponding author: CHRISTIAN HERTEL, Institut fiir Lebensmitteltechnologie, Universitat Hohenheim, Garbenstr. 28, D - 70599 Stuttgart, Germany Phone: +49 711 459 4255; Fax: +49 711 4594199 e-mail: [email protected]