FEMS Microbiology Letters 189 (2000) 19^24
www.fems-microbiology.org
A variation of the ampli¢ed-fragment length polymorphism (AFLP)
technique using three restriction endonucleases, and assessment of
the enzyme combination BglII^MfeI for AFLP analysis of
Salmonella enterica subsp. enterica isolates
BjÖrn-Arne Lindstedt
b
a;
*, Even Heir a , Traute Vardund a , Georg Kapperud
a;b
a
National Institute of Public Health, Department of Bacteriology, N-0403 Oslo, Norway
Department of Pharmacology, Microbiology, and Food Hygiene, Norwegian school of Veterinary Sciences, N-0033 Oslo, Norway
Received 12 April 2000 ; received in revised form 24 May 2000; accepted 25 May 2000
Abstract
We have performed amplified-fragment length polymorphism (AFLP) fingerprinting on a collection of Salmonella enterica subsp. enterica
serovar typhimurium strains with a restriction endonuclease combination (BglII and MfeI) that has previously been used successfully for
typing Campylobacter jejuni isolates with high resolution. Additionally, a variation of the AFLP assay in which two rare cutting restriction
enzymes (XbaI and BsrGI) in combination with the frequent cutter (HinP1I) was examined. The BglII and MfeI enzyme combination
offered low resolution for genotyping Salmonella typhimurium isolates and is not recommended for this common serovar. The three-enzyme
combination gave a higher discrimination, and is thus a new alternate way of performing AFLP fingerprinting of S. typhimurium. ß 2000
Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
Keywords : Ampli¢ed-fragment length polymorphism; Capillary electrophoresis ; Salmonella
1. Introduction
The ampli¢ed-fragment length polymorphism (AFLP)
genotyping method has gained considerable interest, since
it was published by Vos et al. in 1995 [1], as a new and
high resolution ¢ngerprinting method for bacterial species.
AFLP generates ¢ngerprints from DNA of both eukaryotic and prokaryotic origin without any prior knowledge
of the sequence. The method is reviewed extensively by
Janssen et al. and Savelkoul et al. [2,3]. We have previously evaluated the AFLP technique using £uorescently
labeled primers and with capillary electrophoresis separation of the fragments, for Salmonella enterica [4], Escherichia coli (Heir, E., Lindstedt, B.A., Vardund, T., Wasteson, Y. and Kapperud, G., submitted for publication) and
Campylobacter jejuni (Lindstedt, B.A., Heir, E., Vardund,
T., Melby, K.K., and Kapperud, G., submitted for publication). We found that the highest discriminatory power
of AFLP versus pulsed-¢eld gel electrophoresis (PFGE)
* Corresponding author. Tel. : +47 (22) 04 22 00;
Fax: +47 (22) 04 25 18; E-mail : [email protected]
typing was achieved with the enzyme combination BglII^
MfeI used for genotyping C. jejuni (Lindstedt, B.A., Heir,
E., Vardund, T., Melby, K.K., and Kapperud, G. submitted for publication). The EcoRI^MseI restriction endonuclease combination used for typing S. enterica gave the
same level of strain discrimination as PFGE [4]. For Shiga-toxin producing E. coli (STEC) strains, PFGE was
however superior to the EcoRI^MseI AFLP that was
used successfully with S. enterica (Heir, E., Lindstedt,
B.A., Vardund, T., Wasteson, Y. and Kapperud, G., submitted for publication). This led us to examine the use of
the AFLP assay (the BglII^MfeI enzyme combination),
which gave us the best discriminative power versus
PFGE (with C. jejuni), for genotyping S. typhimurium.
In our previous study with AFLP performed on Salmonella strains [4], we additionally used the XbaI restriction
enzyme to increase the discriminatory power of AFLP.
XbaI in combination with MseI resulted in a ¢ngerprint
pattern with a limited number of bands, around 10^15
which is few for typical AFLP generated ¢ngerprints,
and increasing the number of bands would likely give a
higher resolution. In this study we examined a new threeenzyme approach to the AFLP technique using XbaI in
0378-1097 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 1 0 9 7 ( 0 0 ) 0 0 2 4 5 - 7
FEMSLE 9472 14-7-00
20
B.-A. Lindstedt et al. / FEMS Microbiology Letters 189 (2000) 19^24
combination with another rare cutter with a related target
sequence (BsrGI) and a frequent cutter (HinP1I) in order
to increase the number of bands in an AFLP assay with
XbaI.
2. Materials and methods
2.1. Bacterial strains
In all 27 strains of S. enterica subsp. enterica with the
majority of strains (20) belonging to the serovar typhimurium were used in this study. All strains used were obtained from the strain collection at the National Reference
Laboratory for Enteropathogenic Bacteria at the National
Institute of Public Health, Oslo, Norway, that serotypes
bacterial enteropathogens isolated by microbiological laboratories throughout Norway.
2.2. DNA extraction
LB medium was inoculated with pure cultures from
agar plates and incubated overnight at 37³C with gentle
agitation. Genomic DNA was then extracted using a commercial kit (`Easy-DNA' ; Invitrogen BV, Leek, The Netherlands).
2.3. AFLP ¢ngerprinting
500 ng of genomic DNA was incubated at 37³C for 5 h
in a 40-Wl solution containing 1U`One Phor All' bu¡er
(Pharmacia, Uppsala, Sweden) with 4 U of two rare cutting enzymes (BsrGI and XbaI) and 4 U of a frequent
cutting enzyme (HinP1I) (New England Biolabs, Beverly,
MA, USA) and 50 ng Wl31 BSA. After 5 h a 10-Wl ligation
mix was added containing 5 pmol of BsrGI and XbaI
adapters, 50 pmol HinP1I adapter, 1 mM ATP, 1 U T4DNA ligase (New England Biolabs) and 50 ng Wl31 BSA
in 1U`One Phor All' bu¡er (Pharmacia). Incubation was
continued overnight at 16³C. The next day 50 Wl of TE
bu¡er was added, to make the PCR template solution. The
BsrGI adapters were: 5P-CTG GTA GAC TGC GTA CC3P and 5P-CTG ACG CAT GGC ATG-3P. The XbaI
adapters were: 5P-CTA GCG TAC GCA GTC-3P and
5P-CTC GTA GAC TGC GTA CG-3P. The HinP1I adapters were : 5P-CAC GAT GAG TCC TGA A-3P and 5PCTA CTC AGG ACT TGC-3P. One microliter of the
PCR-template solution was used in a 20-Wl PCR mix containing 10 pmol of primers for rare cutting enzymes
(BsrGI and XbaI), 10 pmol of primer for the frequent
cutting enzyme (HinP1I), 2 mM of each dNTP, 0.4 U
Taq polymerase (Sigma, St. Louis, MO, USA) in 1UTaq
bu¡er supplied with enzyme. The PCR reaction was carried out on a Perkin-Elmer GeneAmp PCR system 9700
(Perkin-Elmer Inc., Norwalk, CT, USA). The temperature
pro¢le was: 95³C denaturation for 5 min followed by
10 cycles of 94³C for 30 s, 60³C for 30 s, 72³C for 45 s.
Then 30 cycles of 94³C for 30 s, 56³C for 30 s, 72³C for
1 min and ¢nally a 5-min extension step at 72³C. The
BsrGI primer was 5P-GAC TGC GTA CCG TAC AG3P with 5-prime FAM (5P-carboxy£uorescein) label. The
XbaI primer was 5P-GAC TGC GTA CGC TAG A-3P
with 5-prime FAM label, and the HinP1I primer was unlabeled with the sequence 5P-GAT GAG TCC TGA ACG
C-3P (MWG-Biotech AG, Germany). The BglII and MfeI
assay with primers and adapters were as previously described [5,6].
2.4. Fragment analysis
After PCR ampli¢cation the PCR product was diluted
1:10 with sterile water and 1 Wl was mixed with 12 Wl of
deionized formamide and 0.5 Wl of GeneScan TAMRA500 size standard (PE Biosystems, Foster City, CA, USA).
The mix was denaturated at 95³C for 5 min. After cooling
on ice the samples were subjected to capillary electrophoresis using an ABI-310 Genetic Analyzer with POP-4 polymer (PE Biosystems) at 60³C for 30 min. Preprocessing of
the data was performed using the GeneScan fragment
analysis software (PE Biosystems), and the data were subsequently transported to the GelCompar II software (Applied Maths BVBA, Kortrijk, Belgium) for further analysis. A phylogenetic tree was constructed with GelCompar
II using Dice coe¤cients and cluster analysis with the
unweighted pair group method with arithmetic averages
(UPGMA) from the ABI trace ¢les.
3. Results and discussion
3.1. Three-enzyme AFLP
A large study of AFLP genotyping with EcoRI and
MseI on S. enterica serovars has previously been performed in our laboratory [4]. This study showed that the
AFLP method agreed well with typing by DNA macrorestriction with XbaI and PFGE separation. Some of the
¢ngerprint patterns were however quite di¤cult to distinguish, and in some instances we had to use a di¡erent rare
cutter enzyme (XbaI) to separate patterns that could be
distinguished by PFGE. The rational for including the rare
cutting DNA endonuclease XbaI in an AFLP assay was
that this enzyme has been used successfully in several
PFGE genotyping protocols for Salmonella and E. coli
showing high discriminatory power and high reproducibility [7^11]. However, AFLP genotyping performed with
XbaI combined with the frequent cutter MseI or HinP1I
gave few bands, which could be used for analysis, as compared with the use of EcoRI as the rare cutter. In the
present study, we examined whether the addition of another rare cutter, giving an AFLP reaction with two rare
cutter restriction enzymes that cut Salmonella DNA less
FEMSLE 9472 14-7-00
B.-A. Lindstedt et al. / FEMS Microbiology Letters 189 (2000) 19^24
Fig. 1. AFLP ¢ngerprint of a S. typhimurium isolate using the three restriction endonucleases XbaI, BsrGI and HinP1I. Vertical scale is relative £uorescence and horizontal scale is fragment sizes in bp.
frequent than EcoRI, and one frequent cutter, would give
a higher number of bands in the 50^500-bp size range than
the XbaI^MseI and XbaI^HinP1I combinations. The
BsrGI enzyme was chosen as the second rare cutting
restriction endonuclease because of its target sequence
21
resemblance with XbaI (XbaI; 5P-TCTAGA, BsrGI ;
5P-TGTACA). The enzymes BsrGI and XbaI were combined with the frequent cutter enzymes HinP1I or MseI.
AFLP with three enzymes gave a ¢ngerprint pattern with
sharp and easily distinguishable peaks after capillary electrophoresis (Fig. 1). The highest discrimination was obtained using HinP1I as the frequent cutter and therefore
only these results will be presented. The three-enzyme
combination did increase the number of fragments from
about 14 when XbaI^MseI was used, to over 20 (Fig. 1).
Two of the strains were subjected to three repeated DNA
extractions and three repeated runs, and identical ¢ngerprint pro¢les were displayed for the major bands ( s 500
relative £uorescence). We did observe some variations between repeated runs of the same strain on the height of the
fragment peaks, and also some variations on the number
of the lower intensity peaks (below 500 relative £uorescence). The assay was however stable when only peaks
above 500 relative £uorescence were included in the analysis. Since both the BsrGI and the XbaI primers were
labeled with FAM (5P-carboxy£uorescein), a greater variation in peak £uorescence was expected compared to
AFLP assays with one labeled primer. A higher variability
Fig. 2. Dendrogram of 27 S. enterica isolates made from the three-enzyme AFLP assay. In all, 21 di¡erent pro¢les were resolved when isolates showing
90% or more homology were designated as being identical.
FEMSLE 9472 14-7-00
22
B.-A. Lindstedt et al. / FEMS Microbiology Letters 189 (2000) 19^24
Fig. 3. Dendrogram of 20 S. typhimurium isolates made from the three-enzyme AFLP assay. In all, 16 di¡erent pro¢les were resolved when isolates
showing 90% or more homology were designated as being identical.
in relative £uorescence was indeed observed without causing any di¤culties with the analysis of the ¢ngerprints.
Fig. 2 presents a dendrogram made from the XbaI^
BsrGI^HinP1I AFLP patterns of 27 S. enterica strains.
In all 21 di¡erent pro¢les could be distinguished when a
similarity index of 90% homology was used to de¢ne identical strains. The 20 serovar typhimurium strains gave rise
to 16 di¡erent AFLP pro¢les (Fig. 3). This indicates that
the three-enzyme AFLP combination is a useful approach
for genotyping S. typhimurium isolates.
3.2. AFLP using BglII and MfeI for genotyping Salmonella
We recently performed AFLP genotyping on C. jejuni
subsp. jejuni isolates using the restriction enzymes BglII
and MfeI (Lindstedt, B.A., Heir, E., Vardund, T., Melby,
K.K., and Kapperud, G., submitted for publication). This
enzyme combination gave a higher discriminatory power
for typing Campylobacter than both PFGE and restriction
fragment length polymorphism analysis of polymerase
chain reaction products (PCR-RFLP) of the £aA and
Fig. 4. Dendrogram of 16 S. typhimurium isolates made from the BglII^MfeI AFLP assay. All isolates were more than 95% similar, and gave rise to
only one ¢ngerprint pro¢le.
FEMSLE 9472 14-7-00
B.-A. Lindstedt et al. / FEMS Microbiology Letters 189 (2000) 19^24
23
Fig. 5. Dendrogram of 16 S. typhimurium isolates made from the three-enzyme AFLP assay. In all, 12 di¡erent pro¢les were resolved when isolates
showing 90% or more homology were designated as being identical.
£aB genes (Lindstedt, B.A., Heir, E., Vardund, T., Melby,
K.K., and Kapperud, G., submitted for publication). In
the present study we used exactly the same AFLP assay
that was used, with good result, for Campylobacter. We
however found that when the BglII and MfeI enzyme
combination was used for genotyping Salmonella strains,
only minor variations in the band patterns were observed
resulting in low discrimination. Fig. 4 shows a dendrogram of 16 S. enterica serovar typhimurium isolates typed
with BglII and MfeI. None of the strains displayed di¡erent genotyping pro¢les, or could be distinguished, at the
90% similarity level (i.e. strains showing v 90% similarity
are regarded as identical). When the same 16 Salmonella
serovar typhimurium strains were typed with XbaI, BsrGI
and HinP1I, 12 distinct pro¢les were resolved at the 90%
similarity level (Fig. 5). The three-enzyme based AFLP
assay thus o¡ered superior resolution as compared with
the BglII^MfeI assay. The BglII^MfeI restriction enzyme
combination is thus not recommended for AFLP typing of
Salmonella serovar typhimurium.
3.3. Conclusion
The results presented in this study clearly demonstrate
that using an AFLP assay that showed excellent results on
one bacterial species (C. jejuni) does not necessarily give
good results when migrated to a di¡erent species (S. enterica). We have previously seen that the BglII and MfeI
enzyme combination gave somewhat better discrimination
for genotyping STEC strains, but here the patterns could
not be faithfully reproduced and variations in the fragment patterns were observed on repeated runs of the
same strains (Heir, E., Lindstedt, B.A., Vardund, T.,
Wasteson, Y. and Kapperud, G., submitted for publication). We additionally present in this study a novel varia-
tion of the £uorescent-AFLP genotyping technique which
uses two rare cutting restriction enzymes (BsrGI and
XbaI), both with fewer targets in bacterial DNA than
the widely used EcoRI enzyme, in combination with a
frequent cutter (HinP1I) directed at the GCGC sequence.
This assay was promising both regarding to the sharp and
easy recognizable patterns produced and the number of
distinct pro¢les obtained. No extra time or labor was required by the use of three restriction endonucleases in
AFLP as compared with using two. This study indicates
that considerable e¡ort must be used to evaluate di¡erent
AFLP techniques for every new species under study, both
regarding reproducibility and discriminative power.
References
[1] Vos, P., Hogers, R., Bleeker, M., Reijans, M., van-de-Lee, T.,
Hornes, M., Frijters, A., Pot, J., Peleman, J., Kuiper, M. and Zabeau, M. (1995) AFLP: a new technique for DNA ¢ngerprinting.
Nucleic Acids Res. 23, 4407^4414.
[2] Janssen, P., Coopman, R., Huys, G., Swings, J., Bleeker, M., Vos, P.,
Zabeau, M. and Kersters, K. (1996) Evaluation of the DNA ¢ngerprinting method AFLP as an new tool in bacterial taxonomy. Microbiology 142, 1881^1893.
[3] Savelkoul, P.H., Aarts, H.J., de Haas, J., Dijkshoorn, L., Duim, B.,
Otsen, M., Rademaker, J.L., Schouls, L. and Lenstra, J.A. (1999)
Ampli¢ed-fragment length polymorphism analysis: the state of an
Art. J. Clin. Microbiol. 37, 3083^3091.
[4] Lindstedt, B.A., Heir, E., Vardund, T. and Kapperud, G. (2000)
Fluorescent ampli¢ed-fragment length polymorphism genotyping of
Salmonella enterica subsp. enterica serovars and comparison with
pulsed-¢eld gel electrophoresis typing. J. Clin. Microbiol. 38, 1623^
1627.
[5] Kokotovic, B., Friis, N.F., Jensen, J.S. and Ahrens, P. (1999) Ampli¢ed-fragment length polymorphism ¢ngerprinting of mycoplasma
species. J. Clin. Microbiol. 37, 3300^3307.
[6] Kokotovic, B. and On, S.L. (1999) High-resolution genomic ¢nger-
FEMSLE 9472 14-7-00
24
B.-A. Lindstedt et al. / FEMS Microbiology Letters 189 (2000) 19^24
printing of Campylobacter jejuni and Campylobacter coli by analysis
of ampli¢ed fragment length polymorphisms. FEMS Microbiol. Lett.
173, 77^84.
[7] Puohiniemi, R., Heiskanen, T. and Siitonen, A. (1997) Molecular
epidemiology of two international sprout-borne Salmonella outbreaks. J. Clin. Microbiol. 35, 2487^2491.
[8] Van-Beneden, C.A., Keene, W.E., Strang, R.A., Werker, D.H., King,
A.S., Mahon, B., Hedberg, K., Bell, A., Kelly, M.T., Balan, V.K.,
Mac, K.W. and Fleming, D. (1999) Multinational outbreak of Salmonella enterica serotype Newport infections due to contaminated
alfalfa sprouts. JAMA 281, 158^162.
[9] Wegener, H.C. and Baggesen, D.L. (1996) Investigation of an out-
break of human salmonellosis caused by Salmonella enterica ssp.
enterica serovar Infantis by use of pulsed ¢eld gel electrophoresis.
Int. J. Food Microbiol. 32, 125^131.
[10] Rice, D.H., McMenamin, K.M., Pritchett, L.C., Hancock, D.D. and
Besser, T.E. (1999) Genetic subtyping of Escherichia coli O157 isolates from 41 Paci¢c Northwest USA cattle farms. Epidemiol. Infect.
122, 479^484.
[11] Kariuki, S., Gilks, C., Kimari, J., Obanda, A., Muyodi, J., Waiyaki,
P. and Hart, C.A. (1999) Genotype analysis of Escherichia coli strains
isolated from children and chickens living in close contact. Appl.
Environ. Microbiol. 65, 472^476.
FEMSLE 9472 14-7-00