www.elsevier.com/locate/vetmic

Veterinary Microbiology 109 (2005) 29–36

The diversity of Mycoplasma hyopneumoniae within and

between herds using pulsed-field gel electrophoresis

Tim Stakenborg a,*, Jo Vicca b, Patrick Butaye a, Dominiek Maes b,Johan Peeters a, Aart de Kruif b, Freddy Haesebrouck b

a Veterinary and Agrochemical Research Centre, Groeselenberg 99, 1180 Brussels, Belgiumb Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, 9820 Merelbeke, Belgium

Received 23 December 2004; received in revised form 9 May 2005; accepted 12 May 2005

Abstract

Over the years, pulsed-field gel electrophoresis (PFGE) has been proven a robust technique to type isolates with a high

resolution and a good reproducibility. In this study, a PFGE protocol is described for the typing of Mycoplasma hyopneumoniae

isolates. The potential of this technique was demonstrated by comparing M. hyopneumoniae isolates obtained from the same as

well as from different herds. The use of two different restriction enzymes, SalI and ApaI, was evaluated. For each enzyme, the

resulting restriction profiles were clustered using the unweighted pair group method with arithmetic means (UPGMA). For both

obtained dendrograms, the included isolates of the related M. flocculare species clustered separately from all M. hyopneumoniae

isolates, forming the root of the dendrograms. The PFGE patterns of the M. hyopneumoniae isolates of different herds were

highly diverse and clustered differently in both dendrograms, illustrated by a Pearson’s correlation coefficient of only 0.33. A

much higher similarity was observed with isolates originating from different pigs of a same herd. The PFGE patterns of these

isolates always clustered according to their herd and this for both dendrograms. In conclusion, the results indicate a closer

relationship of M. hyopneumoniae isolates within a herd compared to isolates from different herds and this for both restriction

enzymes used. Since the described PFGE technique was shown to be highly discriminative and reproducible, it will be a helpful

tool to further elucidate the epidemiology of M. hyopneumoniae.

# 2005 Elsevier B.V. All rights reserved.

Keywords: PFGE; M. hyopneumoniae; Molecular epidemiology; Typing

1. Introduction

Respiratory diseases are of major concern for pig

herds all over the world. Typically, Mycoplasma

hyopneumoniae plays an essential role and makes

* Corresponding author. Tel.: +32 2 3790437; fax: +32 2 3790690.

E-mail address: tista@var.fgov.be (T. Stakenborg).

0378-1135/$ – see front matter # 2005 Elsevier B.V. All rights reserved

doi:10.1016/j.vetmic.2005.05.005

the host more vulnerable to infections with

secondary pathogens (Done, 1991). Depending on

the herd, the symptoms may remain subclinical or

steer towards a severe porcine respiratory disease

complex. Herd management and housing conditions

are crucial (Maes et al., 1996), but also the virulence

of the isolate is not to be neglected (Vicca et al.,

2003).

.

T. Stakenborg et al. / Veterinary Microbiology 109 (2005) 29–3630

Apart from virulence, differences between M.

hyopneumoniae isolates were already demonstrated at

antigenic level (Ro and Ross, 1983). At least in part,

these differences are the result of an isolate-specific

post-translational cleavage, as was shown for the P97

adhesin (Djordjevic et al., 2004). Also at genomic

level, M. hyopneumoniae isolates turn out to be very

heterogeneous. A remarkably high variety of isolates

was observed using AFLP (Kokotovic et al., 1999),

RAPD (Artiushin and Minion, 1996), field inversion

gel electrophoresis (Frey et al., 1992) or sequence

analysis of single genes (Wilton et al., 1998). Further

information on the typing of M. hyopneumoniae is

very sparsely available and epidemiological data on

the spreading of the disease are mainly obtained by

clinical observations, the detection of serum anti-

bodies or demonstration of the organism by nested

PCR on nasal swabs (Goodwin, 1985; Hege et al.,

2002; Vicca et al., 2002). Direct contact with infected

animals was shown to be a major risk factor (Morris

et al., 1995), but also transmission by air from other

herds or transport vehicles can (re)infect herds

originally free of M. hyopneumoniae (Goodwin,

1985; Hege et al., 2002; Rautiainen and Wallgren,

2001). Despite these studies, the routes of infection are

not always clear (Maes et al., 2000) and the spreading

of individual clones has not been examined in detail

owing to the limited number of isolates available and

the difficulty to standardize currently described

molecular typing techniques. RAPD generally lacks

interlaboratory reproducibility (Penner et al., 1993),

while the high number of fragments generated during

AFLP, usually of different intensities, complicates

data processing (Hong and Chuah, 2003). Multi-locus

sequence typing was shown to be a highly discrimi-

native and reproducible technique, but has not been

described for M. hyopneumoniae and is still too

expensive for small or medium-sized laboratories

(Olive and Bean, 1999). Therefore, pulsed-field gel

electrophoresis (PFGE), also with high discriminatory

power and interlaboratory reproducibility, remains a

method of choice for the typing of many bacteria

(Tenover et al., 1995; van Belkum et al., 1998).

Although most commonly used to monitor outbreaks,

PFGE also allows to examine chronic infections in

order to better understand transmission patterns

(Struelens et al., 2001). Therefore, in this study, a

PFGE protocol was optimised and used to compare

M. hyopneumoniae isolates obtained within a herd as

well as from different herds.

2. Materials and methods

2.1. Strains and growth conditions

The J-reference strain (National Collection of Type

Cultures (NCTC) 10110), the USA 232 reference strain

(Minion et al., 2004), two Danish field isolates and a

total of 35 M. hyopneumoniae isolates, originating from

21 different Belgian and two different Lithuanian herds,

were used (Figs. 1 and 2). For both Lithuanian and for

eight Belgian herds, isolates from two to three different

pigs within the same herd were included. The isolates

are indicated using the following format: ‘F1.2A’,

where F1 represents the number of the herd, 2 indicates

the number of the pig and A is an arbitrary letter

representing the strain. Isolates originating from

Lithuania received the prefix LH, the Danish isolates

the prefix DK.

The 232 reference strain was received from the

College of Veterinary Medicine (Iowa State University,

USA), while the Danish strains were kindly provided

by the Danish Veterinary Institute (Copenhagen,

Denmark). All Belgian and Lithuanian field strains

were isolated from lungs of pigs at slaughter with

typical M. hyopneumoniae lesions and positive during

immunofluorescence (Kobisch et al., 1978). The

isolation was performed in broth medium according

to Friis (1975) and the identity of the isolates was

confirmed by means of a multiplex-PCR (Stakenborg

et al., in press). For PFGE analysis, the isolates were

cultivated in 40 ml Friis’ medium (Kobisch and Friis,

1996) at 37 8C for at least five days to the end of the

exponential or beginning of the stationary growth

phase. The virulence of eight isolates has been

determined in experimentally inoculated pigs. Isolates

F7.2C and DK Mp143 were of high virulence, isolate

F12.6A was moderately virulent and isolates F1.12A,

F5.6A, F9.8K, the J-strain and F13.7B were of low

virulence (Vicca et al., 2003). The M. flocculare Ms42

reference strain (NCTC 10143) and five Belgian M.

flocculare field isolates were also included and served

as an outgroup during clustering. The Salmonella

enterica serovar Braenderup reference strain H9812

was used as a size marker as proposed by PulseNet

T. Stakenborg et al. / Veterinary Microbiology 109 (2005) 29–36 31

Fig. 1. PFGE patterns of chromosomal DNA of M. hyopneumoniae and M. flocculare isolates restricted with ApaI. Cluster analysis was

performed with UPGMA using the Dice coefficient and a tolerance and optimisation level of 0.8%. Bands below 18 kbp were omitted for

analysis.

(Swaminathan et al., 2001; Hunter et al., 2005) and was

grown overnight at 37 8C on Columbia agar with 5%

ovine blood (Oxoid, UK).

2.2. PFGE

The isolates were harvested by centrifugation at

3000 � g for 15 min. The supernatant was placed in a

new sterile Falcon tube (BD Biosciences, NJ, USA)

and centrifuged a second time using the same

conditions. Both pellets were pooled in 2 ml washing

buffer (50 mM Tris–HCl, 10 mM EDTA, 25% (w/v)

glucose, pH 7.3) and centrifuged at 13000 � g for

5 min. The washed pellets were resuspended in 800 ml

resuspension buffer (75 mM NaCl, 25 mM EDTA, pH

7.3) and the optical density at 610 nm (OD610) was

determined. The bacterial suspension was adjusted to

an OD610 of 1.8 and 200 ml of this suspension was

T. Stakenborg et al. / Veterinary Microbiology 109 (2005) 29–3632

Fig. 2. PFGE patterns of chromosomal DNA of M. hyopneumoniae and M. flocculare isolates restricted with SalI. Cluster analysis was performed

with UPGMA using the Dice coefficient and a tolerance and optimisation level of 0.8%. Bands below 18 kbp were omitted for analysis.

mixed with an equal volume of 1% Seakem Gold agar

(Cambrex Bio Science, ME, USA) at 56 8C and

poured into Plexiglas molds (Bio-Rad, CA, USA) to

set into blocks (5 mm � 2 mm � 10 mm). The blocks

were hardened at 4 8C during 10 min followed by lysis

of the mycoplasma cells using 2 ml freshly prepared

lysis buffer (50 mM EDTA, 1% N-lauroyl-sarcoside,

0.1 mg/ml proteınase K, 10 mM Tris–HCl, pH 8.0) for

18 h at 50 8C. Afterwards, the agarose blocks were

washed three times during 15 min with distilled water,

followed by three washing steps using sterile washing

buffer (50 mM Tris–HCl, 10 mM EDTA, pH 7.3).

Next, plugs were equilibrated during 15 min in 1�restriction buffer (delivered with the enzyme).

Subsequently, the DNA in the plugs was digested

using restriction buffer containing 30 units of ApaI

(Roche, Switzerland) or SalI (MBI Fermentas,

Lithuania) during 4 h at 37 8C. Before electrophoresis,

the plugs were rinsed with Tris-borate–EDTA (TBE

0.5�, 45 mM Tris-borate, 1 mM EDTA, pH 8.0) and

T. Stakenborg et al. / Veterinary Microbiology 109 (2005) 29–36 33

loaded in a 1% Seakem Gold agarose (Cambrex Bio

Science). Electrophoresis was performed for 18 h

under a constant temperature of 14 8C at 6 V/cm and

with a linear switch time rampage from 0.5 to 8.5 s

(CHEF Mapper, Bio-Rad). Salmonella Braenderup

plugs were prepared by the same protocol, but were

restricted with XbaI (Roche). Agarose gels were

stained with ethidium bromide and after destaining in

water for 30 min, the DNA fragments were visualized

using a Genegenius gel documentation system

(Westburg, The Netherlands).

2.3. Data analysis and clustering

The digital images were imported in the Bionu-

merics software (V3.5, Applied Maths, Belgium) and

bands were marked after standardization using the

Salmonella Braenderup restriction fragments. Calcu-

lation of similarity coefficients was performed using

the Dice algorithm. The unweighted pair group

method with arithmetic mean (UPGMA) was used

for clustering. In order to attain a complete match

between strains analysed in duplicate, the band

position tolerance and optimisation were set to

0.8% and bands smaller than 18 kbp were omitted.

The observed PFGE patterns of strain 232 were

compared with the fragments determined in silico

based on its genome sequence (Minion et al., 2004).

The Pearson’s correlation coefficient was calculated

by comparing the Dice similarity coefficient matrices of

both restriction enzymes. In addition to the dendrograms

obtained for both restriction enzymes separately, a

cluster analysis on the average of both separate

dendrograms was calculated using BioNumerics, giving

both independent analyses the same equal weight.

The typeability of the PFGE technique for both

restriction enzymes was determined. To calculate the

discriminatory power, the Simpson’s index was used

(Hunter and Gaston, 1988) with and without including

multiple M. hyopneumoniae isolates originating from

a single farm.

3. Results

The described PFGE protocol resulted in clear

restriction fragments ranging from 18 to 250 kbp

(SalI) or 300 kbp (ApaI). Nicely separated bands were

obtained for the M. hyopneumoniae and the M.

flocculare isolates, for both restriction enzymes used

(Figs. 1 and 2). The Salmonella Braenderup strain

suited perfectly as a marker since well separated

bands, spanning the entire size-range, were obvious

after restriction with XbaI using the same protocol

(data not shown). For one M. hyopneumoniae isolate,

F20.1G, no profile could be obtained after restriction

with ApaI, resulting in a typeability of 97%, compared

to 100% for restriction with SalI. A clear, apparent

band was visible on top of the gel (data not shown),

representing the unrestricted genomic DNA. All other

restriction patterns were clustered for each restriction

enzyme using the UPGMA algorithm.

For strain 232 most restriction fragments deter-

mined in silico were observed on gel as well, although

three fragments differed in size. After restriction with

ApaI, the calculated band of 190 kbp appeared larger

on gel, while the calculated band of 171 kbp was

considerably smaller. After restriction with SalI, the in

silico determined band of 90 kbp was only about half

its size on gel.

The Simpson’s index of diversity gave for both

restriction enzymes a discrimination index of 0.997

provided that isolates from the same farm were

considered related and were not taken into account.

When all M. hyopneumoniae isolates were included,

the discrimination index was still as high as 0.990 for

SalI and 0.983 for ApaI.

All M. flocculare isolates included in this study

clustered together, separately from the M. hyopneu-

moniae isolates (lower than 40% similarity). Only

when six or more isolates were used, the M. flocculare

PFGE patterns formed the root of the tree (see Figs. 1

and 2). Whenever less M. flocculare PFGE patterns

were used, they clustered together, but in-between the

M. hyopneumoniae isolates (data not shown).

A high variety between the PFGE patterns of M.

hyopneumoniae isolates, originating from different

herds, was observed for both restriction enzymes used.

Only the isolates from herd 21 and 23 showed

identical profiles. With the exception of these latter

two isolates and F18.2A and F4.2C after restriction

with ApaI, isolates derived from different farms

showed less than 80% similarity. Clustering of the

highly diverse PFGE patterns generated largely

different dendrograms for both restriction enzymes

used. In other words, isolates showing a high

T. Stakenborg et al. / Veterinary Microbiology 109 (2005) 29–3634

similarity based on ApaI results, may differ largely for

the SalI restriction patterns, and vice versa. A weak,

but still positive, association between the similarity

coefficients for SalI and ApaI was calculated

(Pearson’s correlation coefficient = 0.33).

For the nine isolates tested in an experimental

infection model, no linkage between virulence and

PFGE patterns was observed. Neither did isolates of

the same geographical origin cluster together.

Conversely, PFGE patterns of isolates that were

obtained from different pigs originating from the same

herd clustered together. This was the case for all

isolates analysed and was apparent in both dendro-

grams. With the exception of isolates of herd 19 for

restriction with SalI and Lithuanian herd 3 after

restriction with ApaI, isolates derived from the same

farm showed over 80% similarity. The isolates of herd

15, herd 17 and Lithuanian herd 1 had identical PFGE

profiles for both enzymes used. Isolates of herd 11, 14

and 16 on the other hand had identical profiles for one

restriction enzyme, but small differences were

observed using the second restriction enzyme. The

multiple isolates of the other herds showed small

differences in their PFGE profiles for both enzymes

used.

4. Discussion

The validity of PFGE for molecular typing is well

established (Maule, 1998; Struelens et al., 2001) and

its high discriminatory power and reproducibility

was also apparent in this study. Moreover, the PFGE

protocol optimised for M. hyopneumoniae was

shown useful for the typing of M. flocculare isolates

as well. This was to be expected since both species

are highly related for both their biochemical

and serological characteristics (Kobisch and Friis,

1996; Stemke et al., 1992). Another porcine

mycoplasma, M. hyorhinis, is less related and initial

tests showed indeed that the PFGE protocol using

SalI or ApaI was not useful for the latter species (data

not shown).

An enormous heterogeneity between the studied M.

hyopneumoniae isolates originating from different

herds was observed. These findings are in agreement

with earlier findings obtained by RAPD (Artiushin and

Minion, 1996) and AFLP (Kokotovic et al., 1999). On

the other hand, PFGE patterns of M. hyopneumoniae

isolates originating from a single herd showed more

similarity compared to isolates from different herds.

Many strains from the same herd showed even

identical PFGE patterns, while for the other isolates

small differences were observed. These differences,

together with the high heterogeneity of strains in

general, indicate significant genome plasticity. This is

further substantiated by the comparison of PFGE

results and in silico generated data for strain 232.

These observed differences can be explained by a

single chromosomal inversion event (from 93–96 to

357–361 kbp). Also for the J-reference strain, geno-

mic differences have been reported after in vitro

passages (Frey et al., 1992). On the other hand, the

similarity of PFGE patterns of strains originating from

a single farm does not automatically imply that these

isolates are related, since PFGE is not suited to depict

phylogenetic trees (Davis et al., 2003; Struelens et al.,

2001; Tenover et al., 1995). A recent report on PFGE

of Escherichia coli isolates concluded that in the

absence of other data, six or more restriction patterns

may be needed to estimate the relatedness of isolates

(Davis et al., 2003). This is in agreement with our

calculation of the Pearson’s correlation coefficient,

which was indeed very low. However, both enzymes

show the same trend, namely isolates from the same

herd cluster together. Also, combining the two

enzymes in one single cluster-analysis, showed similar

results (data not shown). Although all isolates were

obtained from slaughter pigs and no relation to the age

of the pig can be made, these results strongly suggest

that isolates from a single herd are derived from only

one or a few ancestral clones.

It is still not known whether these observed genomic

differences are linked to phenotypical differences.

Many reports already demonstrated isolate-dependent

antigenic variations in Mycoplasma species (Bhugra

et al., 1995; Noormohammadi et al., 1997; Rosengarten

et al., 1993; Roske et al., 2001) and also for M.

hyopneumoniae, differences in surface antigens (Wise

and Kim, 1987) and lipid content (Chen et al., 1992)

have been reported. If our PFGE results are indeed

linked to differences on the proteonomic level, these

data may explain why vaccination, although normally

beneficial, often leads to an incomplete protection that

may vary between different herds (Morrow et al., 1994;

Maes et al., 1999).

T. Stakenborg et al. / Veterinary Microbiology 109 (2005) 29–36 35

Although PFGE patterns were similar within a herd,

an enormous variety between isolates was even visible

for isolates originating from a limited geographical

region. These observations are in contrast with an

earlier report where five Swiss strains seemed more

homogenous than five from other origins (Frey et al.,

1992). Probably, more clones of different countries

need to be investigated before definite conclusions can

be made. The same report suggested a possible link

between field inversion gel electrophoresis patterns and

virulence (Frey et al., 1992), but again this could not be

confirmed with our data.

5. Conclusion

In conclusion, the PFGE profiles of M. hyopneumo-

niae isolates originating from different herds were very

diverse, compared to the limited heterogeneity seen

within a herd. Further research may be needed to sustain

these data and to further elucidate the distribution,

stability and persistence of M. hyopneumoniae clones.

The proposed PFGE protocol proved a very useful and

reproducible tool to perform these studies.

Acknowledgements

This work was supported by a grant of the Federal

Service of Public Health, Food Chain Safety and

Environment (Grant number S-6136).

The authors thank Veronique Collet and Sara

Tistaert for skilful technical assistance and Annelies

Pil for the numerous inspiring discussions.

References

Artiushin, S., Minion, F.C., 1996. Arbitrarily primed PCR analysis

of Mycoplasma hyopneumoniae field isolates demonstrates

genetic heterogeneity. Int. J. Syst. Bacteriol. 46, 324–328.

Bhugra, B., Voelker, L.L., Zou, N., Yu, H., Dybvig, K., 1995.

Mechanism of antigenic variation in Mycoplasma pulmonis:

interwoven, site-specific DNA inversions. Mol. Microbiol. 18,

703–714.

Chen, J.W., Zhang, L., Song, J., Hwang, F., Dong, Q., Liu, J., Qian,

Y., 1992. Comparative analysis of glycoprotein and glycolipid

composition of virulent and avirulent strain membranes of

Mycoplasma hyopneumoniae. Curr. Microbiol. 24, 189–192.

Davis, M.A., Hancock, D.D., Besser, T.E., Call, D.R., 2003. Evalua-

tion of pulsed-field gel electrophoresis as a tool for determining

the degree of genetic relatedness between isolates of Escherichia

coli O157:H7. J. Clin. Microbiol. 41, 1843–1849.

Djordjevic, S.P., Cordwell, S.J., Djordjevic, M.A., Wilton, J.,

Minion, F.C., 2004. Proteolytic processing of the Mycoplasma

hyopneumoniae cilium adhesin. Infect. Immun. 72, 2791–2802.

Done, S.H., 1991. Environmental factors affecting the severity of

pneumonia in pigs. Vet. Rec. 128, 582–586.

Frey, J., Haldimann, A., Nicolet, J., 1992. Chromosomal hetero-

geneity of various Mycoplasma hyopneumoniae field isolates.

Int. J. Syst. Bacteriol. 42, 275–280.

Friis, N.F., 1975. Some recommendations concerning primary iso-

lation of Mycoplasma suipneumoniae and Mycoplasma floccu-

lare a survey. Nord. Vet. Med. 27, 337–339.

Goodwin, R.F., 1985. Apparent reinfection of enzootic-pneumonia-

free pig herds: search for possible causes. Vet. Rec. 116, 690–

694.

Hege, R., Zimmermann, W., Scheidegger, R., Stark, K.D., 2002.

Incidence of reinfections with Mycoplasma hyopneumoniae and

Actinobacillus pleuropneumoniae in pig herds located in respira-

tory-disease-free regions of Switzerland—identification and

quantification of risk factors. Acta Vet. Scand. 43, 145–156.

Hong, Y., Chuah, A., 2003. A format for databasing and comparison

of AFLP fingerprint profiles. BMC Bioinform. 4, 7.

Hunter, P.R., Gaston, M.A., 1988. Numerical index of the discri-

minatory ability of typing systems: an application of Simpson’s

index of diversity. J. Clin. Microbiol. 26, 2465–2466.

Hunter, S.B., Vauterin, P., Lambert-Fair, M.A., Van Duyne, M.S.,

Kubota, K., Graves, L., Wrigley, D., Barrett, T., Ribot, E., 2005.

Establishment of a universal size standard strain for use with the

PulseNet standardized pulsed-field gel electrophoresis proto-

cols: converting the national databases to the new size standard.

J. Clin. Microbiol. 43, 1045–1050.

Kobisch, M., Tillon, J., Vannier, P., Magneur, S., Morvan, P., 1978.

Pneumonie enzootique a Mycoplasma suipneumoniae chez le

porc: diagnostic rapide et recherches d’anticorps. Rec. Med. Vet.

Alfort. 154, 847–852.

Kobisch, M., Friis, N.F., 1996. Swine mycoplasmoses. Rev. Sci.

Technol. 15, 1569–1605.

Kokotovic, B., Friis, N.F., Jensen, J.S., Ahrens, P., 1999. Amplified-

fragment length polymorphism fingerprinting of Mycoplasma

species. J. Clin. Microbiol. 37, 3300–3307.

Maes, D., Deluyker, H., Verdonck, M., Castryck, F., Miry, C.,

Vrijens, B., Verbeke, W., Viaene, J., de Kruif, A., 1999. Effect

of vaccination against Mycoplasma hyopneumoniae in pig herds

with an all-in/all-out production system. Vaccine 17, 1024–

1034.

Maes, D., Verdonck, M., de Kruif, A., 2000. Enzootische pneumonie

bij varkens. Deel I: De ziekte (in Dutch) Tijdschr. Diergeneeskd.

69, 94–100.

Maes, D., Verdonck, M., Deluyker, H., de Kruif, A., 1996. Enzootic

pneumonia in pigs. Vet. Q. 18, 104–109.

Maule, J., 1998. Pulsed-field gel electrophoresis. Mol. Biotechnol. 9,

107–126.

Minion, F.C., Lefkowitz, E.J., Madsen, M.L., Cleary, B.J., Swartzell,

S.M., Mahairas, G.G., 2004. The genome sequence of Myco-

T. Stakenborg et al. / Veterinary Microbiology 109 (2005) 29–3636

plasma hyopneumoniae strain 232, the agent of swine myco-

plasmosis. J. Bacteriol. 186, 7123–7133.

Morris, C.R., Gardner, I.A., Hietala, S.K., Carpenter, T.E., Ander-

son, R.J., Parker, K.M., 1995. Seroepidemiologic study of

natural transmission of Mycoplasma hyopneumoniae in a swine

herd. Prev. Vet. Med. 21, 323–337.

Morrow, W., Iglesias, G., Stanislaw, C., Stephenson, A., Erickson,

G., 1994. Effect of a mycoplasma vaccine on average daily gain

in swine. J. Swine Health Prod. 2, 13–18.

Noormohammadi, A.H., Markham, P.F., Whithear, K.G., Walker,

I.D., Gurevich, V.A., Ley, D.H., Browning, G.F., 1997. Myco-

plasma synoviae has two distinct phase-variable major mem-

brane antigens, one of which is a putative hemagglutinin. Infect.

Immun. 65, 2542–2547.

Olive, D.M., Bean, P., 1999. Principles and applications of methods

for DNA-based typing of microbial organisms. J. Clin. Micro-

biol. 37, 1661–1669.

Penner, G.A., Bush, A., Wise, R., Kim, W., Domier, L., Kasha, K.,

Laroche, A., Scoles, G., Molnar, S.J., Fedak, G., 1993. Repro-

ducibility of random amplified polymorphic DNA (RAPD)

analysis among laboratories. PCR Methods Appl. 2, 341–345.

Rautiainen, E., Wallgren, P., 2001. Aspects of the transmission of

protection against Mycoplasma hyopneumoniae from sow to

offspring. J. Vet. Med. B. Infect. Dis. Vet. Public Health 48,

55–65.

Ro, L.H., Ross, R.F., 1983. Comparison of Mycoplasma hyopneu-

moniae isolates by serologic methods. Am. J. Vet. Res. 44,

2087–2094.

Rosengarten, R., Theiss, P.M., Yogev, D., Wise, K.S., 1993. Anti-

genic variation in Mycoplasma hyorhinis: increased repertoire of

variable lipoproteins expanding surface diversity and structural

complexity. Infect. Immun. 61, 2224–2228.

Roske, K., Blanchard, A., Chambaud, I., Citti, C., Helbig, J.H.,

Prevost, M.C., Rosengarten, R., Jacobs, E., 2001. Phase variation

among major surface antigens of Mycoplasma penetrans. Infect.

Immun. 69, 7642–7651.

Stakenborg, T., Vicca, J., Butaye, P., Imberechts, H., Peeters, J., de

Kruif, A., Haesebrouck, F., Maes, D. A multiplex PCR to

identify porcine mycoplasmas present in broth cultures. Vet.

Res. Comm., in press.

Stemke, G.W., Laigret, F., Grau, O., Bove, J.M., 1992. Phylogenetic

relationships of three porcine mycoplasmas, Mycoplasma

hyopneumoniae, Mycoplasma flocculare, and Mycoplasma

hyorhinis, and complete 16S rRNA sequence of M. flocculare.

Int. J. Syst. Bacteriol. 42, 220–225.

Struelens, M.J., De Ryck, R., Deplano, A., 2001. Analysis of

microbial genomic macrorestriction patterns by pulsed-

field gel electrophoresis (PFGE) typing. In: Dijkshoorn, L.,

Towner, K.J., Struelens, M. (Eds.), New Approaches for the

Generation and Analysis of Microbial Typing Data. Elsevier

B.V., Amsterdam, pp. 159–176.

Swaminathan, B., Barrett, T.J., Hunter, S.B., Tauxe, R.V., 2001.

PulseNet: the molecular subtyping network for foodborne bac-

terial disease surveillance, United States. Emerg. Infect. Dis. 7,

382–389.

Tenover, F.C., Arbeit, R.D., Goering, R.V., Mickelsen, P.A., Murray,

B.E., Persing, D.H., Swaminathan, B., 1995. Interpreting chro-

mosomal DNA restriction patterns produced by pulsed-field gel

electrophoresis: criteria for bacterial isolate typing. J. Clin.

Microbiol. 33, 2233–2239.

van Belkum, A., van Leeuwen, W., Kaufmann, M.E., Cookson, B.,

Forey, F., Etienne, J., Goering, R., Tenover, F., Steward, C.,

O’Brien, F., Grubb, W., Tassios, P., Legakis, N., Morvan, A., El

Solh, N., de Ryck, R., Struelens, M., Salmenlinna, S., Vuopio-

Varkila, J., Kooistra, M., Talens, A., Witte, W., Verbrugh, H.,

1998. Assessment of resolution and intercenter reproducibility

of results of genotyping Staphylococcus aureus by pulsed-field

gel electrophoresis of SmaI macrorestriction fragments: a multi-

center study. J. Clin. Microbiol. 36, 1653–1659.

Vicca, J., Maes, D., Thermote, L., Peeters, J., Haesebrouck, F., de

Kruif, A., 2002. Patterns of Mycoplasma hyopneumoniae infec-

tions in Belgian farrow-to-finish pig herds with diverging dis-

ease-course. J. Vet. Med. B. Infect. Dis. Vet. Public Health. 49,

349–353.

Vicca, J., Stakenborg, T., Maes, D., Butaye, P., Peeters, J., de Kruif,

A., Haesebrouck, F., 2003. Evaluation of virulence of Myco-

plasma hyopneumoniae field isolates. Vet. Microbiol. 97, 177–

190.

Wilton, J.L., Scarman, A.L., Walker, M.J., Djordjevic, S.P., 1998.

Reiterated repeat region variability in the ciliary adhesin gene of

Mycoplasma hyopneumoniae. Microbiology 144, 1931–1943.

Wise, K.S., Kim, M.F., 1987. Major membrane surface proteins of

Mycoplasma hyopneumoniae selectively modified by covalently

bound lipid. J. Bacteriol. 169, 5546–5555.