The influence of molecular oxygen exposure on the biology of Prevotella intermedia , with emphasis...

ORIGINAL ARTICLE

The influence of molecular oxygen exposure on the biologyof Prevotella intermedia, with emphasis on its antibioticsusceptibilityS.G. dos Santos1, C.G. Diniz2, V.L. da Silva1, N.C. Souza1, F.L. de Lima1, M.R.Q. Bomfim1,M.A.R. de Carvalho1 and L.M. Farias1

1 Departamento de Microbiologia, Instituto de Ciencias Biologicas, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil

2 Departamento de Parasitologia, Microbiologia e Imunologia, Instituto de Ciencias Biologicas, Universidade Federal de Juiz de Fora, Juiz de Fora,

Minas Gerais, Brazil

Introduction

Prevotella intermedia is a rod-shaped, gram-negative,

anaerobic black-pigmented bacterium that belongs to the

Bacteroidaceae family (Shah and Gharbia 1992) and may

be found in several kinds of clinical specimens from

patients with infection in almost all the anatomic sites

(Finegold 1995; Santos et al. 2003).

Anaerobic therapy is based on modification of the

environment to slow down the proliferation of microbes

with the use of appropriate antimicrobial agents. Gener-

ally, an antimicrobial drug is selected according to the

knowledge of suspected pathogens, their anticipated

antimicrobial susceptibility pattern and favourable phar-

macokinetic properties in an environment with low pH

and oxygen tension (Falagas and Siakavellas 2000).

According to the literature, however, studies have docu-

mented an alarming but gradual increase in resistance

rates of anaerobic bacteria to several antibiotics in

hospitals in the United States, Europe and elsewhere

(Falagas and Siakavellas 2000; Aldridge et al. 2001;

Santos et al. 2004).

Keywords

aerotolerance, antimicrobial susceptibility,

Prevotella intermedia.

Correspondence

Luiz de Macedo Farias, Departamento de

Microbiologia, Instituto de Ciencias Biologicas,

Universidade Federal de Minas Gerais.

Avenida Antonio Carlos, 6627, Pampulha;

CEP 31.270-901, Belo Horizonte, MG, Brazil.

E-mail: [email protected]

2006 ⁄ 1487: received 25 October 2006,

revised 18 December 2006 and accepted 4

January 2007

doi:10.1111/j.1365-2672.2007.03313.x

Abstract

Aim: This study focuses on investigating the molecular and physiological char-

acteristics of Prevotella intermedia after molecular oxygen exposure (MOE) and

the effect on drug susceptibility patterns.

Methods and Results: Samples of P. intermedia were used as parent strains:

ATCC25611 and four clinical isolates. Strains adapted to oxidative stress by

MOS were obtained by the enrichment technique. Drug susceptibility was

evaluated by minimal inhibitory concentrations (MIC) using agar dilution.

Arbitrarily primed-polymerase chain reaction (AP-PCR) was used to evaluate

the genetic diversity of all strains and physiological analyses were made by

sodiumdodecylsulfate-polyacrylamide gel electrophoresis and two-dimensional

electrophoresis of crude, cell-free extracts. The genetic profile showed that line-

ages with altered MIC values were selected after MOE. Overall, we found signi-

ficant decrease in drug susceptibility for the aero-strains against all tested

antimicrobials (amoxicillin, amoxicillin + clavulanic acid, clindamycin, chlo-

ramphenicol, ertapenen and metronidazole). We also observed markedly differ-

ent protein expression patterns between the parent and selected aero-strains.

Conclusions: MOE induces changes in the genetic profile and protein expres-

sion patterns of P. intermedia that may also be linked to its drug resistance

mechanisms.

Significance and Impact of the Study: The effects of MOE on anaerobic

bacterial physiology and behaviour may influence antimicrobial susceptibility

patterns with potential consequences to antimicrobial chemotherapy.

Journal of Applied Microbiology ISSN 1364-5072

882 Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 882–891

ª 2007 The Authors

Antibiotics that are frequently used against anaerobic

bacteria include metronidazole, imipenem and merope-

nem which are active against almost all anaerobic bacteria

from clinical isolates, although resistant strains to these

agents have been sporadically reported (Aldridge et al.

2001; Santos et al. 2004; Behra-Miellet et al. 2006). In

contrast to these less common antibiotic resistance pat-

terns, there are two classes that are more commonly seen.

First, resistance to the combination of b-lactam and b-

lactamase inhibitor amoxicillin ⁄ clavulanic acid has been

noted in some multicentre in vitro antimicrobial surveys,

showing somewhat higher resistance rates (Santos et al.

2004; Behra-Miellet et al. 2006). Second, resistance to

clindamycin and cephalosporins such as cefoxitin has

been widely spread and is potentially significant (Teng

et al. 2002). On the other hand, recently developed fluor-

oquinolones such as tosufloxacin, demonstrate improved

antianaerobic activity (Takahata and Nishino 1997).

Because of its cost effectiveness, especially in developing

countries, chloramphenicol continues to be widely used

against anaerobes despite its known haemotoxicity

(Turton et al. 1999).

Oxidative stress has been defined as a disturbance in

the pro-oxidant–antioxidant balance in favour of pro-oxi-

dants (Zheng and Storz 2000). Reactive oxygen species

(ROS) such as O2–, HO and H2O2 are extremely harmful

as they easily penetrate membranes and diffuse through

cells (Zheng and Storz 2000). Thus, the alteration of nor-

mal levels of ROS may cause injury by oxidative stress

raising lesions in different tissues. Several substances can

affect ROS production by cells during oxidative processes

such as exposure to oxygen, radiation, organic peroxides,

metals and chemical substances which may generate intra-

cellular free radicals; a mechanism utilized by some anti-

microbial drugs (Zheng and Storz 2000; Albesa et al.

2004).

The alteration of oxidative metabolism and the increase

of oxygen may affect the mechanism of action of various

antibiotics. Studies indicate that diverse antibiotics can

generate an increase of O)2 in different bacterial species

(Albesa et al. 2004; Koutsolioutsou et al. 2005). Cipro-

floxacin, ceftazidime, piperacillin and chloramphenicol

stimulate oxidative stress in bacteria that may be useful in

explaining aspects of antibacterial action which are, as of

yet, unaccounted (Albesa et al. 2004; Koutsolioutsou et al.

2005).

The production of enzymes involved in oxygen meta-

bolism by an organism – such as superoxide dismutase,

catalase and peroxidase – is mirrored by variable degrees

of tolerance and virulence in some strains and, therefore,

in their pathogenic potential (Silva et al. 2005). The abil-

ity to respond to oxidative stress when exiting in its nat-

ural anaerobic environment to more oxygenated tissues as

well as to resist oxidative events from the host immune

system until anaerobic conditions are restored in the

infection site is an important acquired adaptation for

anaerobic bacteria (Lynch and Kuramitsu 2000).

Little data on the regulation of oxidative stress control

in Prevotella spp. are available. Increasing knowledge

about ROS neutralizing enzymes offers enticing prospects

for improving our understanding in several aspects of the

molecular biology, physiology and ecology of this import-

ant microbial group and how its members participate in

infectious diseases.

Our objective in the present study was to evaluate the

P. intermedia response to oxidative stress by exposing this

anaerobic bacterium to molecular oxygen and then iden-

tify the consequences to microbial biology with regards to

its antimicrobial drug susceptibility patterns.

Materials and methods

Bacterial strains

Five P. intermedia strains were used as parent strains in

this study: the reference P. intermedia ATCC25611 (Wt-

ref) and four P. intermedia strains isolated from patients

with intra-abdominal infections (Wt-34, Wt-71, Wt-96

and Wt-134).

All bacterial strains were routinely cultivated anaerobi-

cally, at 37�C (anaerobic chamber; Forma Scientific Com-

pany, Marietta, OH, USA, containing an atmosphere of

N2 85%, H2 10%, CO2 5%) in Brucella broth (BB; Difco,

Sparks, MD, USA), supplemented with 0Æ05 mg dl)1 of

hemin, 0Æ01 mg dl)1 of menadione, 5% yeast extract

(BB-S) and stored at –86�C (Holdeman et al. 1977).

Throughout the experiments, the strains were checked

by gram staining and aerotolerance tests (Holdeman et al.

1977).

Adaptation of Prevotella intermedia ATCC 25611 and

Prevotella intermedia clinical isolates to atmospheric

molecular oxygen

Strains adapted to oxidative stress were derived from the

parent P. intermedia ATCC25611 (Wt-ref) and the clinical

isolates (Wt-34, Wt-71, Wt-96 and Wt-134) by atmo-

spheric oxygen exposure, as already described (Silva et al.

2005). Briefly, a 300-ml volume of BB was inoculated

with nearly 10% of an overnight culture of each parent

strain and the resulting bacterial suspension was grown to

mid-exponential phase. The bacterial cultures were split

into two 150-ml subcultures. One of the subcultures was

kept in the anaerobic chamber as a control and the other

was placed in an aerobic shaker (Cientec; Biosan, Sao

Paulo, SP, Brazil) adjusted to 200 rev min)1 and 37�C for

S.G. dos Santos et al. Influence of molecular oxygen on Prevotella intermedia biology

ª 2007 The Authors

Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 882–891 883

aeration and temperature. Every 6 h, up to a maximum

of 72 h, samples from both subcultures were withdrawn

and aliquots were evaluated for OD at 600 nm and were

plated after 10-fold serial dilutions, in triplicate, to count

colony-forming units per millilitre (CFU ml)1) onto

Brucella agar supplemented with 0Æ05 mg dl)1 hemin,

0Æ01 mg dl)1 menadione and 5% yeast extract (BA-S). All

plates used to recover bacteria for viable cell counts were

incubated anaerobically, at 37�C, for 48 h. The plates cor-

responding to the last point of atmospheric oxygen expo-

sure at which survival colonies were observed were used

to re-isolate P. intermedia strains to be used for further

adaptive steps. One colony of the surviving derived ones

from each bacterial strain were subcultured in BB for

24 h in an anaerobic chamber and then submitted to four

cycles of 12 h atmospheric oxygen exposure and sub-

culturing as before. The strain resulting from the fourth

re-isolation was considered as the aero-tolerant strain

(Aero-ref, Aero-34, Aero-71, Aero-96 and Aero-134).

Survival curve of Prevotella intermedia strains selected

by molecular oxygen exposure

Two volumes of 600 ml BRU-S each were inoculated

with 10% of an overnight culture (initial OD600 nm =

0Æ136) of WT-strains or aero-strains, respectively, and the

cultures were incubated in the anaerobic chamber at

37�C, until the mid-exponential phase. Then, each culture

was split into two 300-ml subcultures. One 300-ml

aliquot of each culture was kept inside the anaerobic

chamber as control and the other 300-ml aliquot was

removed and kept under aerobic conditions on a shaker

at 200 rev min)1 and 37�C. Every 6 h, up to 144 h, sam-

ples from all subcultures were withdrawn and aliquots

were evaluated for OD at 600 nm and plated onto BRU-S

after 10-fold serial dilutions, in triplicate, for viable cell

counts (CFC ml)1).

Biochemical identification and identity confirmation of

selected strains by oxygen exposure

The identification and the identity confirmation of the

bacterial strains selected after molecular oxygen exposure

(MOS), including the P. intermedia reference strain, were

based on the biochemical, physiological or enzymatic

characteristics using the Rapid-ID 32A system (Bio-

Merieux, Marcy- I’Etoile, France).

Crude cell-free protein extracts from the parent

Prevotella and their derivative strains selected after MOS

Crude cell-free protein extracts were performed according

to Shibata et al. (2005) and (Westeimeier and Naven)

2002. Pellets from mid-log phase cultures of parent (Wt)

and aero-strains were washed in sterile sucrose buffer

(10 mmol l)1 of Tris, 250 mmol l)1 of sucrose, pH 7)

and adjusted to a concentration of approximately 108

cells ml)1 (OD = 0Æ4, at 600 nm). Five-millilitre aliquots

of each cell suspension were precipitated by adding 50%

trichloroacetic acid (Sigma, St. Louis, MO, USA) to a

final concentration of 10% to prevent degradation of

proteins by proteases or endopeptidases and centrifuged

at 8500 g for 10 min at 4�C. The harvested cells

were solubilized in 4 ml of lysis solution [7 mol l)1

of urea, 2 mol l)1 of thiourea, 4% 3-[(3-cholamido-

propyl)-dimethylammonio]-1-propanesulfonate (CHAPS),

40 mmol l)1 of Tris and 1 mmol l)1 of EDTA), and were

lysed by ultrasonication (Branson sonifier 450; Branson

Ultrasonics Corp., Danbury, CT, USA), followed by cen-

trifugation for 15 min at 10 000 g (Laemmli 1970), and

then protein content was measured (PlusOne 2-D Quant

kit; Amersham Biosciences, Little Chalfont, UK).

Identification of differentially regulated proteins by

sodiumdodecylsulfate-polyacrylamide gel electrophoresis

The crude cell-free protein extracts were fractionated in

sodiumdodecylsulfate-polyacrylamide gel electrophoresis

(12% SDS-PAGE) (Laemmli 1970). Electrophoresis was

performed in Tris-glycine-SDS buffer (10 mmol l)1 of

Tris, 250 mmol l)1 of sucrose, pH 7), under constant cur-

rent of 30 mA over 4 h. Proteins were visualized by

Coomassie blue staining (Westeimeier and Naven 2002).

Identification of differentially regulated proteins by

two-dimensional electrophoresis

The crude protein extracts were submitted to two-dimen-

sional gel electrophoresis, as described before (Rocha

et al. 1996; Westeimeier and Naven 2002; Shibata et al.

2005). Briefly, the first dimension electrophoresis, based

on isoelectric focusing, was carried out using a poly-

acrylamide strip, with a pH gradient from 4Æ0 to 7Æ0(IPG; Immobiline DryStrip Kit; Amersham Pharmacia

Biotech), which was loaded with 200 lg of crude cell

extract. Electrophoresis was performed for 20 h, at 20�C,

1 mA and 5 W, using a gradient programming electro-

phoresis power supply EPS 3500 XL (Amersham Pharma-

cia Biotech), as follows: 100 V, 0Æ01 h; 100 V, 4 h; 300 V,

0Æ01 h; 300 V, 4Æ5 h; 2000 V, 5 h and 2000 V, 6Æ5 h

(Veith et al. 2001). With isoelectric focusing complete,

second dimension electrophoresis was carried by 12%

SDS-PAGE (Laemmli 1970). Finally, the gels were stained

with colloidal Coomassie brilliant blue G 250 for analysis

(Westeimeier and Naven 2002). The assays were per-

formed in triplicate.

Influence of molecular oxygen on Prevotella intermedia biology S.G. dos Santos et al.


ª 2007 The Authors

Arbitrarily primed-polymerase chain reaction typing

Arbitrarily primed-polymerase chain reaction (AP-PCR)

typing for the Prevotella strains was based on previous

studies (Kononen et al. 1998; Matto et al. 1999). Briefly,

the isolates were grown on BRU-S plates under anaerobic

atmosphere at 37�C for 24 h. For each parent or oxygen

exposure selected strain, 10 single colonies were collected

from the plates from which DNA was extracted using

Wizard Genomic DNA Purification Kit (Promega Cor-

poration, Madison, WI, USA). Primers Opa-13 (5¢-CAG-

CACCCAC-3¢) and Opa-03 (5¢-AGTCAGCCAC-3¢) were

employed for typing because of better discriminatory

power shown in previous studies (Kononen et al. 1998;

Veith et al. 2001) (Integrated DNA Technologies, Coral-

ville, IA, USA). AP-PCR was performed using the follow-

ing mix (50-ll final volume): 2X PCR Master Mix

(Promega Corporation, Madison, WI, USA), 0Æ8 lmol l)1

of primer and 5 ll of DNA suspension (1 ⁄ 100 dilution).

Amplifications were carried out in a DNA thermal cycler

(PCR Express Thermo Hybaid, Ashford, Middlesex, UK).

An internal negative control was included in each

AP-PCR run. Samples were amplified by initial denatura-

tion at 94�C for 5 min followed by 35 cycles with dena-

turation at 94�C for 1 min, annealing at 32�C for 2 min

and extension at 72�C for 2 min, and the final extension

at 72�C for 5 min (Kononen et al. 1998; Matto et al.

1999).

Analysis of the DNA fragments amplified by AP-PCR

Amplified products were separated by PAGE stained with

silver nitrate (Oliveira et al. 1995) and digitally photo-

graphed. Then, the gels were carefully analysed in order

to determine the main fragments of the AP-PCR profiles

in comparison with the molecular size of the fragments of

a marker, 1-kb DNA Ladder (Promega Corporation). The

similarity among the AP-PCR profiles obtained with the

DNA of the parent or the oxygen exposure-selected strain

P. intermedia was analysed with the LabImage -1D gel

analysis software, version 2Æ7Æ2. Copyright 1999–2004,

Kapelan GmbH, Halle (Saale), Germany. Available at:

http://www.labimage.net (Nei 1972).

Dendrogram

AP-PCR profiles were scored based on the presence (i) or

absence (ii) of each polymorphic DNA fragment gener-

ated by each primer (OPA-3 ⁄ OPA-13) with DNA from

the parent strains and those submitted to oxidative stress.

Cluster analysis based on similarity using Nei’s (1972)

original distance was performed by the Unweighted Pair

Group Method (UPGMA) with arithmetic averages clus-

tering algorithm (Sneath and Sokal 1973) and the rand-

omization procedure as implemented in TFPGA (Tools

for Population Genetic Analyses) software package (Miller

1998).

Minimum inhibitory concentration determination

The minimum inhibitory concentration (MIC) was per-

formed using agar dilution method (Holdeman et al.

1977; NCCLS 2004). Antibiotic stock solutions were

added to melted BA (Difco) to obtain final concentra-

tions ranging from 0Æ05 to 1025Æ0 lg ml)1. The anti-

microbial drugs were selected based on microbial

characteristics and clinical relevance: metronidazole

(Oxoid Ltd, Basingstoke, Hampshire, England), ertape-

nem (Merck & Co., Inc., Whitehouse Station, NJ, USA),

clindamycin (Sigma Chemical Co., St. Louis, MO, USA),

amoxillin (Sigma), choramphenicol (Sigma) and amoxi-

cillin–clavulanate (Oxoid). All samples were grown near

108 UFC ml)1 corresponding to 0Æ5 on the MacFarland

scale, and then inoculated into BB-S (Difco) with antimi-

crobial agents, using a ‘Steers’ replicator (Steers et al.

1959). The reference strains Bacteroides fragilis (ATCC

25285) and Eubacterium lentum (ATCC 43055) were also

included as controls. The plates were incubated in an

anaerobic chamber at 37�C for 48 h. The MIC was meas-

ured as the lowest concentration able to inhibit microbial

growth as interpreted according to Clinical Laboratories

Standards Institute (formerly, National Committee for

Clinical Laboratory Standards) recommendations (NCCLS

2004).

Results

Survival curve of Prevotella intermedia ATCC 25611

(Wt-strain) and aero-strains under atmospheric oxygen

exposure

Five P. intermedia strains adapted to oxidative stress from

atmospheric oxygen exposure were obtained (Aero-ref,

Aero-34, Aero-71, Aero-96 and Aero-134) from parental

strain ATCC 25611 (Wt-ref) and the intra-abdominal

infection clinical isolates (Wt-34, Wt-71, Wt-96 and

Wt-134), after five cycles of aeration, according to the

aforementioned methods. The efficiency of the adaptive

process was demonstrated by the progressive increase in

aerotolerance of P. intermedia strains observed after each

successive cycle. To confirm the adapted phenotype, sur-

vival curves under oxidative conditions were compared

between parent and aero-strains. Viable cells were no lon-

ger observed after 48-h exposure to atmospheric oxygen

for all parent strains, whereas aero-strains were viable for

up to 96 h under the same conditions.


ª 2007 The Authors


Biochemical identification

Biochemical–physiological characterization of P. interme-

dia strains subjected to oxidative stress by MOS did not

show significant alterations. The resulting phenotypes

were similar to that of the reference strain P. intermedia

ATCC 25611.

Comparison between protein expression patterns of parent

Prevotella intermedia and aero-strains selected by MOS

The protein expression patterns were compared for all

tested strains by SDS-PAGE analysis. In addition, the pro-

tein expression patterns of P. intermedia ATCC 25611

(Wt-ref) and its derivative strain (Aero-ref) were also

compared using two-dimensional eletrophoresis and

visualized by Coomassie staining. As shown in Fig. 1, frac-

tioning the crude cell-free protein extracts through two-

dimensional eletrophoresis shows, significant differences

between the parent (Wt) and aero-strains. Both downreg-

ulated and upregulated peptides were observed with

more than 18 differences seen overall by two-dimensional

eletrophoresis. Typical results are shown in Fig. 1, where

the reproducible differences in protein profiles between

the parent (Wt) and the derivative strains that are induced

by exposure to molecular oxygen are highlighted.

Genetic diversity between the parent Prevotella

intermedia and aero-strains selected by MOS

Figure 2 shows the DNA fragments by amplified AP-PCR

from P. intermedia and aero-strains selected by MOS,

using two different primers OPA-3 and OPA-13. The AP-

PCR fragment sizes were determined by the apparent

electrophoretic mobility on polyacrylamide gel using Lab-

Image software (Miller 1998) and ranged from 521 to

4072 bp using primer OPA-3 or 461 to 4072 bp using

primer OPA-13.

The migration profile of the amplified DNA fragments

by the OPA-3 primer was heterogeneous between each

parent and the atmospheric oxygen-selected strains. Con-

sidering Wt-ref and Aero-ref strains, six polymorphic

DNA fragments were detected. When comparing Wt-71

and Wt-134 with their aero-tolerant selected strains four

and two polymorphic DNA fragments were detected

respectively. For the strains Wt-34 and Aero-34, and Wt-

96 and Aero-96, similar patterns were observed with

OPA-3 primer. Regarding primer OPA-13, the strains

Wt-ref ⁄ Aero-ref strains, Wt-34 ⁄ Aero-34 and Wt-71 and

Aero-71 showed more than one polymorphic DNA frag-

ment. The strains Wt-96 and Aero-96 showed one poly-

morphic DNA fragment and the strains Wt-134 ⁄ Aero-134

showed a similar pattern (Tables 1 and 2).

Dendrogram

Figure 3 shows a dendrogram generated by the UPGMA

clustering algorithm that grouped the results of AP-PCR

according to phenotypical similarity among profiles pro-

duced with two arbitrary primers using DNA extracted

from the parent P. intermedia or aero-strains as the tem-

plate. The dendrogram constructed with the results

obtained with two primers used in combination allow

one to cluster the samples into two groups according to

kDa Wt-strain Aero-strain22515010075

50 9

1035

25

15

10

pH 7·0 4·0 pH 7·0 4·0

1 2

34

8

5

11 11

612

7 17

14

18

13

15 1514

13

18 16

177

6 8

549

101

3

2

12

16

Figure 1 Two-dimensional eletrophoresis of cell-free extracts from Prevotella intermedia ATCC 25611 strain (Wt-strain) and from its derivative

strain selected by the oxidative stress by molecular oxygen exposure (aero-strain). Circles and squares indicate proteins that are, respectively,

down- and upregulated in the aero-strain. The figure shows the result of one experiment representative of three independently performed ones.



ª 2007 The Authors

the similarity between them. Group 1 contains the sam-

ples amplified with OPA-13 primer and in group 2 those

samples amplified by AP-PCR using OPA-3 primer.

MIC determinations

We observed that all P. intermedia strains (Wt) tested in

this study had an altered drug susceptibility profile after

MOS (aero-strains, Table 3). Overall, a significant

decrease in drug susceptibility of the aero-strains for all

tested antimicrobials (amoxicillin, amoxicillin + clavul-

anic acid, clindamycin, chloramphenicol, ertapenen and

metronidazole) was observed. Taken together as a mean

value for the aero-strains, if compared with that for the

parent strains, the recorded MIC value was 8Æ89 times

higher for chloramphenicol, 6Æ93 times higher for amoxi-

bp

M 1

PrimerOPA-3

PrimerOPA-3

PrimerOPA-3

PrimerOPA-3

PrimerOPA-3

PrimerOPA-13

PrimerOPA-13

PrimerOPA-13

PrimerOPA-13

PrimerOPA-13

2 1 2 3 4 3 4 5 6 5 6 7 8 7 8 9 10 9 10 M

407230542036

1018

517506

396344298

220202

Figure 2 Electrophoretic profile generated by arbitrarily primed-polymerase chain reaction (AP-PCR) by employing primers OPA-3 and OPA-13 of

Prevotella intermedia parent or oxidative stress-adapted strains. M, Molecular size of marker; lanes 1, 3, 5, 7 and 9, parent strains (ATCC 25611,

34, 71, 96 and 134, respectively); lanes 2, 4, 6, 8 and 10, derivative strains selected by exposure to molecular oxygen (ATCC 25611, 34, 71, 96

and 134, respectively).

Table 1 Relative size (in base pairs) of the specific arbitrarily primed-polymerase chain reaction (AP-PCR) products from Prevotella intermedia

parent or oxidative stress-adapted strains obtained by amplification with OPA-3 and OPA-13 primers, respectively

Tested strains

AP-PCR products*

OPA-3 primer OPA-13 primer

Wt-ref � 521, 586, 747, 1183, 1340, 1656, 1850, 1954, 4072 921, 1057, 1374, 1656, 2464, 4072

Aero-ref� 521, 528, 586, 1183, 1656, 1954, 2240 461, 586, 921, 1057, 1125, 1656, 2036, 2464

Wt-34� 521, 586, 670, 747, 1183, 1340, 1656, 1954, 4072 461, 921, 1057, 1340, 1480, 1656, 2464, 4072

Aero-34� 521, 586, 670, 747, 1183, 1340, 1656, 1954, 4072 461, 921, 1057, 1139, 1480, 1656, 2036, 2464

Wt-71� 521, 586, 747, 1183, 1340, 1656, 1850, 1954, 4072 921, 1340, 1656, 2464, 4072

Aero-71� 521, 586, 1183, 1656, 4072 1139, 1656, 2036, 2464

Wt-96� 521, 586, 670, 747, 1183, 1340, 1656, 1850, 1954, 4072 461, 586, 921, 1057, 1340, 1656, 2464, 4072

Aero-96� 521,586, 670, 747, 1183, 1340, 1656, 1850, 1954, 4072 461, 921, 1057, 1139, 1340, 1656, 2464, 4072

Wt-134� 521, 586, 670, 747, 1183, 1656, 4072 461, 921, 1057, 1139, 1183, 1340, 2036, 2464, 4072

Aero-134� 521, 586, 670, 747, 1183, 1340, 1656, 2036, 4072 461, 921, 1057, 1139, 1183, 1340, 2036, 2464, 4072

Bold numbers represent DNA bands with high intensity, as observed after electrophoresis.

*Used primers by AP-PCR.

�Reference strains of P. intermedia.

�Prevotella intermedia strains submitted to molecular oxygen exposure.

Abbreviations: Pi-Wt, P. intermedia ATCC 25611 wild-type; Pi-Aero, lineage of P. intermedia exposed to molecular oxygen; 34-Wt, 71-Wt, 96-Wt

and 134-Wt, P. intermedia strains obtained from clinical isolates; 34-Aero, 71-Aero, 96-Aero and 134-Aero, lineage of P. intermedia exposed to

molecular oxygen.


ª 2007 The Authors


cillin, 3Æ10 times higher for clindamycin, 4Æ57 times

higher for the combination treatment of amoxicil-

lin + clavulanic acid, 2Æ87 times higher for ertapenen and

2Æ52 times higher for metronidazole. Although there were

decreased sensitivity patterns to the antimicrobial drugs

tested, full resistance was not observed for the tested

aero-strains.

Discussion

Our study reinforces this idea as the obligate anaerobe P.

intermedia was able to develop an efficient response to

MOS that resulted in a survival time at least twice as

long.

Data generated by SDS-PAGE and two-dimensional

electrophoresis show a complex shift in the aero-strains

protein expression patterns when compared with their

parent Wt-strains band profiles as revealed by both

down- and upregulated proteins with different molecular

mass and isoelectric points (Fig. 1). Differences of these

magnitudes may have not resulted from the accumulation

of single point mutations and thus may be part of a

broader programme of genetic control that alters the

physiology to protect and prevent oxidative stress damage

to the cell. However, our results suggest that for these

bacteria the differences observed did not significantly alter

the major metabolic pathways in these cells, once no sig-

nificant alterations were observed in aero-strains based on

the biochemical and physiological or enzymatic character-

istics using the Rapid-ID 32A system.

These global changes, while not entirely understood, do

have other significant effects on the organisms, as related

to their virulence patterns, reflecting the host–bacterial

relationship, as already reported (Silva et al. 2005). Previ-

ous studies document (Rocha et al. 1996; Rocha and

Smith 1997; Fletcher et al. 2001) that environmental fac-

tors, including atmospheric conditions, may induce phy-

siological modulations as observed in this study. Also, the

influence of antimicrobial agents, as another environmen-

tal factor, on the regulation of protein expression by B.

fragilis has already been demonstrated (Diniz et al. 2004).

During entry into the host organism, anaerobic bacteria

may be exposed to high oxygen levels and may be persist-

ently stressed by oxidative bursts that originate from

the host immune response (Katsuragi et al. 2003). It is

assumed that an adaptive response to this oxidative stress

is a key factor for bacterium survival (Rocha et al. 1996).

It seems that this ability represents an important charac-

teristic for an invasive process from the natural anaerobic

environment to more oxygenated tissues of the body

(Lynch and Kuramitsu 2000; Zheng and Storz 2000; Silva

et al. 2005).

The AP-PCR method may be applied to detect poly-

morphisms in a wide variety of organisms using different

primers (George et al. 1997). This technique has been fre-

quently used for clonal analysis, being particularly useful

for determining whether isolates of the same taxa are epi-

demiologically related (Camarena et al. 1995; Song 2005).

In this study, the combination genotyping using two

arbitrary primers, showed lower discriminatory power of

AP-PCR typing with the primers OPA-3 than with OPA-

13 (Fig. 2). Similar results have also been reported with

these same primers and the AP-PCR method when

applied to detect polymorphisms in Neisseria gonorrhoeae

(Camarena et al. 1995).

As observed by the arbitrary amplifications of DNA in

all bacterial strains after MOS by AP-PCR, several altera-

tions were detected in the banding patterns generated

(Tables 1 and 2) and this fact might not be considered

merely a phenotypic adaptive response (Fig. 3). Moreover,

alterations were also observed by the cellular proteins’

electrophoregrams (Fig. 1).

Although there have been great advances in microbial

genomics and proteomics over the past decade, little is

known regarding the mechanisms of bacterial oxidative

stress adaptive response and the influences of this adapta-

tion on the micro-organism–host relationship. In this

regard, it may be useful to evaluate the aerotolerance of

Table 2 Numbers of polymorphic fragments generated by arbitrarily

primed-polymerase chain reaction (AP-PCR) using OPA-3 and OPA-13

primers

Used primers*

OPA-3 OPA-13

Total of

fragments

Polymorphic

fragments

Total of

fragments

Polymorphic

fragments

Wt-ref� 9 4 6 2

Aero-ref c 7 2 8 4

Wt-34� 9 0 8 2

Aero-34� 9 0 8 2

Wt-71� 9 4 5 3

Aero-71� 5 0 4 2

Wt-96� 10 0 8 1

Aero-96� 10 0 8 1

Wt-134� 7 0 9 0

Aero-134� 9 2 9 0

Total 84 12 73 17

*Used primers by AP-PCR.

�Reference strains of Prevotella intermedia.

�Prevotella intermedia strains submitted to molecular oxygen expo-

sure.

Abbreviations: Pi-Wt, P. intermedia ATCC 25611 wild-type; Pi-Aero,

lineage of P. intermedia exposed to molecular oxygen; 34-Wt, 71-Wt,

96-Wt and 134-Wt, P. intermedia strains obtained from clinical

isolates; 34-Aero, 71-Aero, 96-Aero and 134-Aero, lineage of P. inter-

media exposed to molecular oxygen.



ª 2007 The Authors

the bacterium in conditions that mimic those found in

the host, in order to help identify key changes associated

with oxidative stress resistance.

Studies have indicated that antimicrobial drugs are able

to generate ROS in different bacteria (Albesa et al. 2004;

Koutsolioutsou et al. 2005). Several antibiotics used in

medicine kill bacteria by inducing oxidative stress. Low

antibiotic concentrations induce antioxidant systems and

results in increased resistance of the bacteria, which leads

to increasing therapeutic doses of antibiotics. In addition,

Table 3 Antimicrobial susceptibility patterns, by the agar dilution technique, of parent Prevotella intermedia strains (Wt) and their derivative

strains selected after oxidative stress by molecular oxygen exposure (Aero) to antimicrobial drugs of clinical and microbiological relevance

Antimicrobials

ATCC* Strain 34� Strain 71� Strain 96� Strain 134�

Wt Aero Wt Aero Wt Aero Wt Aero Wt Aero

Amoxicillin 0Æ12 0Æ48 0Æ48 2Æ00 0Æ24 2Æ00 0Æ48 4Æ00 0Æ48 4Æ00

Amoxicillin + clavulanic acid 0Æ12 0Æ48 0Æ48 1Æ00 0Æ12 1Æ00 0Æ24 2Æ00 0Æ24 1Æ00

Clindamycin 0Æ48 2Æ00 4Æ00 8Æ00 4Æ00 8Æ00 0Æ48 8Æ00 2Æ00 8Æ00

Chloramphenicol 0Æ12 2Æ00 0Æ48 4Æ00 0Æ24 4Æ00 0Æ48 4Æ00 0Æ48 2Æ00

Ertapenen 0Æ12 0Æ48 0Æ24 1Æ00 0Æ24 0Æ48 0Æ48 2Æ00 1Æ00 2Æ00

Metronidazole 0Æ12 0Æ12 0Æ48 1Æ00 0Æ24 0Æ24 1Æ00 4Æ00 0Æ48 0Æ48

Drug breakpoints according to the NCCLS (2004) recommendations: amoxicillin (2 lg ml)1), amoxicillin + clavulanic acid (16 ⁄ 8 lg ml)1), clinda-

mycin (8 lg ml)1), chloramphenicol (32 lg ml)1), ertapenen (16 lg ml)1) and metronidazole (32 lg ml)1). Results are expressed as minimal inhib-

itory concentrations (MIC – lg ml)1).

*Prevotella intermedia – reference strain ATCC 25611.

�Prevotella intermedia – strains isolated from patients with intra-abdominal infection.

8000 6000 4000 2000 0000

71-Aero

71-Aero34-Aero

Pi-Wt

34-WtGroup 1OPA-13

Group 2OPA-3

96-Aero96-Wt

134-Aero134-Wt

71-Wt

Pi-Aero

Pi-Wt

134-Aero

34-Aero

96-Aero

Pi-Aero

96-Wt

34-Wt

134-Wt

71-Wt

60% 70% 80% 90% 100%

Figure 3 Dendrogram of the Unweighted Pair Group Method (UPGMA) cluster analysis resulting from the comparison of the electrophoretic pat-

terns obtained by arbitrarily primed-polymerase chain reaction (AP-PCR) with OPA-3 and OPA-13 primers from Prevotella intermedia parent and

their derivative strains submitted to oxidative stress by molecular oxygen exposure. Scale of similarity with Nei’s (1972) original distance with boot-

strap of 1000 replicates. Pi-Wt, P. intermedia ATCC 25611 wild-type; Pi-Aero, lineage of P. intermedia exposed to molecular oxygen; 34-Wt,

71-Wt, 96-Wt and 134-Wt, parent P. intermedia strains obtained from clinical isolates. 34-Aero, 71-Aero, 96-Aero and 134-Aero, derivative

strains of P. intermedia exposed to molecular oxygen.


ª 2007 The Authors


the increase in antioxidant potential could generate

new bacterial strains resistant to antibiotics, particularly

those that affect bacteria via the generation of ROS

(Koutsolioutsou et al. 2001).

The strains selected in this study after MOS showed

altered drug susceptibility profiles with significant decrea-

ses, up to four times, in drug sensitivity to clindamycin

followed by chloramphenicol and amoxicillin, compared

with the parent strains. This altered susceptibility might

be related to the microbial protective effect against oxida-

tive stress as they were selected by previous oxygen expo-

sure. However, as MIC values did not achieve the break

points of the assayed drugs, antimicrobial resistance was

not observed for any strain. In general, the phenomenon

was markedly observed for clindamycin followed by chlo-

ramphenicol and amoxicillin.

As suggested before and according to our results, it is

possible that MOS may induce an upregulation in the

protective enzymes of P. intermedia that may be also

linked with drug resistance mechanisms (Koutsolioutsou

et al. 2001, 2005; Albesa et al. 2004).

However, there is still much to be learned regarding

the precise mechanisms of the adaptive response to MOS

and how they might evolve inside the host. This is a pre-

liminary observation and further prospective studies are

needed to better understand how oxidative stress may

influence the drug resistance phenomenon and conse-

quently antimicrobial chemotherapy, concerning anaer-

obic infections.

Acknowledgements

The authors are grateful to Luzia Rosa Rezende and Jose

Sergio de Barros (AT ⁄ CNPq) for their technical support

and to the institutions Conselho Nacional de Desenvolvi-

mento Cientıfico e Tecnologico (CNPq), Fundacao de

Amparo a Pesquisa do Estado de Minas Gerais (FAP-

EMIG) and Coordenacao de Aperfeicoamento de Pessoal

de Nıvel Superior (CAPES) for their financial support

and facilities.

References

Albesa, I., Becerra, C.M., Battan, P.C. and Paez, L.P. (2004)

Oxidative stress involved in the antibacterial action of dif-

ferent antibiotics. Biochem Biophys Res Commun 317, 605–

609.

Aldridge, K.E., Asheraft, D., Cambre, K., Pierson, C.L., Jenkins,

S.G. and Rosenblatt, J.E. (2001) Multicenter survey of the

changing in vitro antimicrobial susceptibility of isolates of

Bacteroides fragilis group, Prevotella, Fusobacterium, Por-

phyromonas, and Peptostreptococcus species. Antimicrob

Agents Chemother 45, 1283–1284.

Behra-Miellet, J., Dubreuil, L. and Calvet, L. (2006) Evaluation

of the in vitro activity of ertapenem and nine other com-

parator agents against 337 anaerobic bacteria. Int J Anti-

microb Agents 28, 25–35.

Camarena, J.J., Nogueira, J.M., Dasi, M.A., Moreno, F., Garcia,

R., Ledesma, E., Lorca, J. and Hernandez, J. (1995) DNA

amplification fingerprinting for subtyping Neisseria gonor-

rhoeae strains. Sex Transm Dis 22, 128–136.

Diniz, C.G., Farias, L.M., Carvalho, M.A.R., Rocha, E.R. and

Smith, C.J. (2004) Differential gene expression in a Bacter-

oides fragilis metronidazole-resistant mutant. J Antimicrob

Chemother 54, 100–108.

Falagas, E.M. and Siakavellas, S. (2000) Bacteroides, Prevotella,

and Porphyromonas species: a review of antibiotic resistance

and therapeutic options. Int J Antimicrob Agents 15, 1–9.

Finegold, S.M. (1995) Overview of clinically important anaer-

obes. Clin Infect Dis 20, 205–207.

Fletcher, J.M., Nair, S.P., Ward, J.M., Henderson, B. and

Wilson, M. (2001) Analysis of the effect of changing envi-

ronmental conditions on the expression patterns of expor-

ted surface-associated proteins of the oral pathogen

Actinobacillus actinomycetemcomitans. Microb Pathog 30,

359–368.

George, K.S., Reynolds, M.A. and Falkler, W.A. Jr (1997) Arbi-

trarily primed polymerase chain reaction fingerprinting

and clonal analysis of oral Fusobacterium nucleatum iso-

lates. Oral Microbiol Immunol 12, 219–226.

Holdeman, L.V., Catto, E.P. and Moore, W.E. (1977) Anaer-

obic Laboratory Manual, 4th edn. Virginia, USA: Institute

and State University Anaerobe Laboratory.

Katsuragi, H., Ohtake, M., Kurasawa, I. and Saito, K. (2003)

Intracellular production and extracellular release of oxygen

radicals by PMNs and oxidative stress on PMNs during

phagocytosis of periodontopathic bacteria. Odontology 91,

13–18.

Kononen, E.J., Matto, M.L., Vaisanen-Tunkelrott, M.L., Frand-

sen, E.V.G., Helander, I., Asikainen, S., Finegold, S.M. and

Jousimies-Somer, H. (1998) Biochemical and genetic char-

acterization of a Prevotella intermedia ⁄ nigrescens-like

organism. Int J Syst Bacteriol 48, 39–46.

Koutsolioutsou, A., Pena-Llopis, S. and Demple, B. (2005)

Constitutive soxR mutations contribute to multiple-antibi-

otic resistance in clinical Escherichia coli isolates. Antimic-

rob Agents Chemother 49, 2746–2752.

Koutsolioutsou, A., Martins, E.A., White, D.G., Levy, S.B. and

Demple, B. (2001) A soxRS-constitutive mutation contri-

buting to antibiotic resistance in a clinical isolate of Sal-

monella enterica (Serovar Typhimurium). Antimicrob

Agents Chemother 45, 38–43.

Laemmli, U.K. (1970) Cleavage of structural proteins during

the assembly of the head of bacteriophage T4. Nature 227,

680–685.

Lynch, M. and Kuramitsu, H. (2000) Expression and role of

superoxide dismutases (SOD) in pathogenic bacteria.

Microbes Infect 2, 1245–1255.



ª 2007 The Authors

Matto, J., Asikainen, S., Vaisanen, M.-L., Von Troil-Linden, B.,

Kononen, E., Saarela, M., Salminen, K., Finegold, S.M., et al.

(1999) b-lactamase production in Prevotella intermedia,

Prevotella nigrescens, and Prevotella pallens genotypes and

in vitro susceptibilities to selected antimicrobial agents.

Antimicrob Agents Chemother 43, 2383–2388.

Miller, M.P. (1998) TFPGA: Tools for Population Genetic Ana-

lyses for Windows. Arizona State University, USA. <http://

www.public.asu.edu/~mmille8/>. Accessed on 15 July

2001.

National Committee for Clinical Laboratory Standards

(NCCLS) (2004) Methods for antimicrobial susceptibility

testing of anaerobic bacteria. Approved Standard – 6th edn.

M11-A6. Wayne, Pennsylvania.

Nei, M. (1972) Genetic distance between populations. Am Nat

106, 283–292.

Oliveira, M.A., Caballero, O.L., Neto, E.D., Koury, M.C.,

Romanha, A.J., Carvalho, J., Hartskeerl, R.A. and Simpson,

A.J. (1995) Use of nondenaturing silver-stained polyacryla-

mide gel analysis of polymerase chain reaction amplifica-

tion products for the differential diagnosis of Leptospira

interrogans infection. Diagn Microbiol Infect Dis 22, 343–

348.

Rocha, E.R. and Smith, C.J. (1997) Regulation of Bacteroides

fragilis katB mRNA by oxidative stress and carbon limita-

tion. J Bacteriol 179, 7033–7039.

Rocha, E.R., Selby, T., Coleman, J.P. and Smith, C.J. (1996)

Oxidative stress response in anaerobe, Bacteroides fragilis: a

role for catalase in protection against hydrogen peroxide.

J Bacteriol 178, 6895–6903.

Santos, S.G., Serufo, J.C., Silva, R.A.P, Marra, B.A., Reis,

C.M.F., Hamdan, J.S., Nicoli, J.R., Carvalho, M.A.R., et al.

(2003) Microbiologic profile of intra-abdominal infections

at Belo Horizonte, Brazil. Am J Infect Control 136, 135–

142.

Santos, S.G., Carvalho, M.A.R., Serufo, J.C., Pinto-Silva, R.A.,

Albuquerque, W., Rosa, R.V., Pires, A.F.B., Hahn, R.C.,

et al. (2004) Antimicrobial susceptibility of microorgan-

isms recovered from intraabdominal infections at Belo

Horizonte, Brazil. Am J Infect Control 32, 414–416.

Shah, H.N. and Gharbia, S.E. (1992) Biochemical and chemical

studies on strains designated Prevotella intermedia and pro-

posal of a new pigmented species Prevotella nigrescens sp.

nov. Int J Syst Bacteriol 42, 542–546.

Shibata, Y., Nakano, Y., Kawada, M., Kojima, M., Fukamachi,

S., Okano, S., Matsushira, K., Abiko, Y., et al. (2005) A

searcheable database for proteomes of oral microorgan-

isms. Oral Microbial Immunol 20, 344–348.

Silva, V.L., Diniz, C.G., Cara, D.C., Santos, S.G., Nicoli, J.R.,

Carvalho, M.A.R. and Farias, L.M. (2005) Enhanced

pathogenicity of Fusobacterium nucleatum adapted to oxi-

dative stress. Microbe Pathogen 39, 131–138.

Sneath, P.H.A. and Sokal, R.R. (1973) Numerical taxonomy:

The Principles and Practice of Numerical Classification.

San Francisco: Freeman & Co.

Song, Y. (2005) PCR-based diagnostics for anaerobic infec-

tions. Anaerobe 11, 79–91.

Steers, E., Foltz, E.L. and Graves, B.S. (1959) An inocula repli-

cating apparatus for routine testing of bacterial susceptibil-

ity to antibiotics. Antibiot Chemother 9, 307–311.

Takahata, M. and Nishino, T. (1997) Antibacterial activities of

tosufloxacin against anaerobic bacteria and the electron

micrograph of its bactericidal effects. Chemotherapy 43,

153–158.

Teng, L.J., Hsueh, P.R., Tsai, J.C., Liaw, S.J. and Luh, K.T.

(2002) High incidence of cefoxitin and clindamycin resist-

ance in Taiwan. Antimicrob Agents Chemother 46, 2908–

2913.

Turton, J.A., Yallop, D., Andrews, C.M., Fagg, R., York, M.

and William, T.C. (1999) Haemotoxicity of chlorampheni-

col succinate in the CD-1 mouse and Wistar Hanover rat.

Hum Exp Toxicol 18, 566–567.

Veith, P.D., Talbo, G.H., Slakeski, N. and Reynolds, E.C.R.

(2001) Identification of a novel heterodimeric outer mem-

brane protein of Porphyromonas gingivalis by two-dimen-

sional gel electrophoresis and peptide mass fingerprinting.

Eur J Biochem 268, 4748–4757.

Westeimeier, R. and Naven, T. (2002) Proteomics in Practice.

Weinheim, Germany: Wiley-VCH Verlag-GmbH.

Zheng, M. and Storz, G. (2000) Mechanisms for sensing and

responding to oxygen deprivation. In Bacterial Stress

Responses ed.Storz, G. and Hengge-Aronis, R. pp. 47–59.

Washington, DC: American Society for Microbiology.


ª 2007 The Authors


The influence of molecular oxygen exposure on the biology of Prevotella intermedia , with emphasis...

Documents

Transcript of The influence of molecular oxygen exposure on the biology of Prevotella intermedia , with emphasis...