Studies on the Oxidative Stress
Response of
P orphyromonus gingiv alis Studies on the Oxidative Stress Response
of P orphyromonas gingivalis
A thesis submitted in fulfillment of the requirements for
admission to the degree of Doctor of Philosophy
By
Patricia I. Díaz
BDS (CES, Colombia), BScDent (Hons)
Dental School
The University of Adelaide
South Australia
December,2002
Acknowledgments
I wish to thank a number of people who provided me with invaluable assistance during the course of this work.
Assoc. Prof. Tony Rogers, who supervised my studies, for his encouragement and advice.
Dr. Neville Gully, for his advice and technical assistance on many occasions and for reading critically this manuscript.
Mr. Peter Z\lm, for his advice and technical assistance on many occasions.
Dr. Renato Morona, Department of Molecular and Life Sciences, for his advice and for allowing me the use of the facilities in his laboratory. I also want to acknowledge his staff and students for their technical advice.
Ms. Lyn Waterhouse, Centre for Electron Microscopy and Microstructure
Analysis, for her technical advice on SEM-related techniques.
Dr. Phil Burcham and Dr. Frank Fontaine, Department of Clinical and
Experimental Pharmacology, for allowing me the use of some facilities in their laboratory.
Dr. Nada Slakeski, School of Dental Sciences, The University of Melbourne, for providing me with invaluable advice and allowing me to spend some time in her laboratory. I also want to thank her staff who provided me with technical support during the course of my experiments in Melbourne.
Dr. David Parker, for reading this manuscript.
My family and friends, for their understanding, faith and endurance.
The University of Adelaide, for awarding nìe an Iuternational Postgladuate
Research Scholarship.
tv Parts of the present studies were supported by grants from The Australian Dental
Research Foundation Inc Summary
Porphyromonas gingivalis is a Gram-negative anaerobic cocco-bacillus
strongly implicated in the aetiology of adult periodontitis. During the colonisation
of the oral cavity it is likely that P. gingivalis encounters different sources of
oxidative stress. Adaptation to this challenge is necessary for the microorganism to
survive and establish in the periodontal environment. The aim of the present study
was, therefore, to investigate the oxidative stress survival mechanisms of P.
gingivalis.
P. gingivalls was grown under different oxygenated environments in a
continuous culture system under conditions of haemin-limitation and excess.
Steady state was achieved under moderately oxygenated atmospheres, although a
decrease in cell viability was observed as the oxygen concentration in the gas
mixture increased. Haemin-excess conditions seemed to increase the ability of a
culture to cope with a determined oxygen concentration. The main change in
fermentation end-products characterising oxygen stressed cultures was an increase
in the production of acetate. Scanning electron micrographs showed that oxygen
triggers a change in the cell shape from a cocco-bacillary to a short rod.
The effect of oxygen on the expression of cysteine proteinases, critical
virulence factors of P. gingivalis, was assayed in supernatants and cell fractions
and further analysed by 2-D gel electrophoresis. Both evaluations showed an
increase in the cell-associated Arg-gingipain and a decrease in Lys-gingipain in
oxygen stressed cells.
Assays for NADH oxidase, NADH peroxidase and superoxide dismutase in
cell extracts showed an increase in their activities as the environment became more
VI oxygenated. The NADH oxidase activity was partially purified and characterised and, surprisingly, the isolated protein was identified as a 4-hydroxybutyryl-CoA dehydratase. This is the first report of NADH oxidase activity associated with this enzyme.
The existence of other open reading frames in the P. gingivalls genome sequence that would encode for proteins with the potential for NADH oxidase/peroxidase activity was further investigated. The transcription product of the identified ORFs, encoding for a possible NADH oxidase (Nox) and an alkyl hydroperoxide reductase (AhpF-C), was analysed under anaerobic and oxygenated environments by northern blot hybridization. No mRNA for Nox was detected but
AhpF-C was expressed constitutively in anaerobic cells and slightly increased under oxygenated conditions.
Furthermore, the possibility of the existence of a common transcriptional switch for oxidative stress-related proteins was investigated. A homologue of
OxyR, a redox-sensitive transcriptional activator, was identified in the P. gittgivalis genome sequence and an OxyR-disrupted mutant was constructed. Mutants exhibited decreased tolerance to air and hydrogen peroxide and were characterised by the absence of alkyl hydroperoxide reductase mRNA, suggesting a control of
OxyR over its expression.
The second part of this research project consisted of studies of P. gingivalis in co-culture with F. nucleatum, vnder oxygenated environments. These studies showed that not only does F. nucleatum have a much higher tolerance to oxygen than P. gingivalis but, in co-culture, it can protect the latter organism and increase its ability to survive under oxygenated conditions. Additionally, it was observed
vll that F. nucleatum is able to generate a capnophilic environment essential for the growth of P. gingivalis.
In conclusion, this study showed that P. gingivalis possesses basic mechanisms to cope with moderate or transient oxidative stress while it probably relies on the protection of other organisms, like F. nucleatum, to survive and replicate in highly oxygenated atmospheres.
vlll Contents
ACKNOWLEDGMENTS IV
SUMMARY VI
1. TNTRODUCTTON...... 1
1.1 PERTODONTAL DTSEASES ...... 2
1.2 PonpavnoMoNAs G\NG\VALIS...... 3
1.2.1 Generalconsideratiot?s...... J
1.2.2 Nutritional requirements and c),totoxic end-product fonnation...... 5
1.2.3 Proteinases ...... ó
1.2.4 Interactions of P. girtgivalis with epitheLial cells and other nticroorgan.istns...... l0
1.3 REGULATION oFGENE EXPRESStoN tn P. GtNctvl,¿/s...... l2
I .4 OxyceN TOXICITY To ANAEROBES AND MICROBIAL RESPONSE TO OXYCEN .,...... ,...... ,...... ,.... 15
1.4.1 Ox¡,gen toxicit¡' to anaerobes t5
I . 4. 2 A nt i - o x idan t e n zy nte.ç ...... l8
I .4.3 Global regulation of tlte anti-oxidant response ...... z-)
I.5 OXYGEN AS A ENVIRONMENTAL CUE IN THE ORAL ENVIRONMENT AND THE OXIDATIVE STRESS
RESPONSE OF P. C1NG|VAL\5...... 25
1.6 Sunvrnnv AND AIMS 3t
2. CHARACTERISATION OF THE GROWTH OF P. GINGIVALIS UNDER
OXYGENATED ENVIRONMENTS 34
2.I EFFECToFEXPOSURETOOXYGENONTHEVIABILITYOFINDIVIDUALCELI-S...,..... 35
2.1.1 Methods 35
2.L2 Results 36
,39
2.2.1 Metlrcds 39
IX 2.2.1 .1 Microorganism and maintenance of the strain...... 39
2.2.1 .2 Continuous culture growth conditions...... 39
2.2.1.3 Statistical Analysis...... 40
2.2.2 Re s u Lt s and disct ts s io,? ...... 41
2,3 EFFECT oF HAEMIN oN THE CONTINUOUS CULTURE CROWTH AND OXYCEN TOLERANCE ...... 44
2.3.1 Detenn.ination of tlte optitttum lutemin concentration to be used.. 45
2.3.1.1 Methods
2.3.1.2 Results ...... 45
2.3.2 Effect of ox¡,gen on the haentin-limited continuous culture growtlt of P. gittgivaLis...... 46
2.3.2.1Methods 46
2.3.2.2 Results and discussion...... 41
2.4 MORPHOLOGICAL CHANGES IN P. jINGIVA¿IS GROWN UNDER ANAEROBIC AND OXYCENATED
ENVIRONMENTS ....
2.4.1 SEM anaLysig 49
2.4.1.1 Methods ...... 49
2.4.1.2 Results
2.4.2 Anal:,sis of capsuLe production... 5l
2.4.2.1 Methods ...... 51
2.4.2.2 Results 5l
2,5 EFFECT oF OXYGEN ON THE FORMATION OF ACIDIC METABOLIC END-PRODUCTS 5l
2.5.I Methods 51
2.5.2 Results... 52
2.6 SUMMARY OF RESULTS AND DISCUSSION 56
3. EFFECT OF OXYGEN ON P. GINGIVA¿IS CYSTEINE PROTEINASES
62
3.1 EFFECT oF OXYGEN ON THE ACTIVITY OF ARC- AND LYS-PROTEINASES 63
3. I . I M ct hod.r ...... 63
3. L2 Res ult s and discus siorl...... 64
X 3.2 ANALYSTS OF PROTEINASES IN P. GINGIVALIS OUTER MEMBRANES BY 2-D CEL
ELECTROPHORESIS.....,...... 68
3.2.I Methods 68
3.2.1 .1 Growth and harvest of P. gingivalis 68
3.2.1.2 Outer membrane preparation 68
3.2.1.3 2D PAGE...... 69
3 .2.1 .4 In-gel digestion of proteins separated by 2D PAGE ...... 70
3.2.1 .5 Peptide mass fin gerprinting (PMF)...... 7l
3.2.1.6 Protein identification 7l
3.2.2 Results and discusston 72
3.2.2.1 Results for Arg-proteinase (RgpA and RgpB) 13
3.2.2.2 Results for Lys-proteinase (Kgp)...... t-)
3.3 Surr¿runny oF RESULTS AND DlscussloN....
4. STUDIES ON ENZYMES INVOLVED IN THE METABOLISM OF
OXYGEN BY P. GINGIVA¿IS...... 80
4.1 THe EFFECT oF oxYGEN oN THE ACTIVITY oF ANTI-oxIDANT ENZYMES 81
4.1.1 Methods 81
4.1.1 .1 Preparation of crude cell extracts.. 8l
4.1 .1 .2 Enzyme Assays. .... 81
4.1 .1 .3 Further analysis of NADH oxidase activity...... 82
4. L2 Re suh s and discus si on ...... 8-i
4.1 .2.1 Effect of 02 on the activity of anti-oxidant enzymes of P. gingivalls grown under
haemin-excess and haemin-limi ted conditions. 83
4.1 .2.2 Furthel analysis of NADH oxidase activity...... 85
4.2 ISOLATION AND CHARACTERISATION OF NADH OXIDASE ACTIVITY....., ...... '.89
4.2.1 Mcthods
4.2.1 .1 Determination of the effect of the cysteine proteinase inhibitor TLCK on NADH
oxidase activity 89
XI 4.2.1 .2 Growth conditions and harvesting of cells...... 90
4.2.1 .3 Preparation of call extracts...... 90
4.2.1.4 Protein content and enzyme activity determinations. 90
4.2.1 .5 Ammonium sulphate fractionation 9t
4.2.1.6 Ion Exchange Chromatography (IEC)
4.2.1.1 Hydrophobic Interaction Chromatography (HIC) ,.'.'.,.,.'.',.,.,.,,.,,', 92
4.2.1.8 Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS PAGE) ...... 92
4.2.1 .9 Molecular mass estimatlon...... 93
4.2.1 .10 Analysis of sample 54...... 93
4.2.1 .l I Identification of the protein with NADH oxidase activity in sample 54...... 94
4.2.1 .12 Mass spectrometry (MS) identification of purified protein...... 95
4.2.2 Results and discussrcn
4.2.2.1 Determination of the effect of the cysteine proteinase inhibitor TLCK on NADH
oxidase activity
4.2.2.2 NADH oxidase purification ...... 96
4.2.2.3Identification of the protein with NADH oxidase activity in sample 54...... 100
4.2.2.4 Mass spectrometry identification t0r
4.2.2.5 Analysis of NADH oxidase activity in sample 54
4.3 AN¡.r-vsrs oFPG 0625 sseupNce......
4.3. I Gene 1D...... I 06
4.3 .1 .l Methods ...... 107
4.3 .l .2 Results .. 108
4.3.2 Furthet' evidence of NADH oxidase activi1t in an AbfD protein ...... 1l6
4.4 IDENTTFICATIoN oF OTHER CANDIDATE CENES RESPONSIBLE FOR NADH OXIDASE ACTIVITY IN
THE P. GINGIVA¿IS CENOME SEQUENCE......
4.4.1 Methods
4.4.2 Results and discussion.....
4.5 EXPRESSION oFNoX AND AHPF-C UNDER ANAEROBIC AND OXYCENATED ENVIRONMENTS I25
125
4.5.1.1 Growth of P. gingivalis V/50......
xll 4.5.1 .2 RNA isolation ...... 125
4.5.1 .3 Determination of RNA purity and concentlation 126
4.5.1.4 Electrophoresis and blotting of RNA...... 126
4.5.1 .5 Probe preparation...... 127
4.5.1.6 Northern hybridization and detection 129
4.5.2 Results 129
4.5.2.1 Expression of AhpF under anaerobic and oxygenated environments 129
4.5.2.2 Expression of Nox under anaerobic and oxygenated environments...... 130
4.6 SUMMARY oF RESULTS AND DISCUSSION...... ,,,.. 13t
5. STUDIES ON POSSIBLE REGULATORY SYSTEMS OF THE
OXIDATIVE STRBSS RESPONSE IN P. GINGIVALIS 1,4r
5,I IDENTTFTCATION oF POSSIBLE TRANSCRIPTIONAL RECULATORS OF THE OXIDATIVE STRESS
RESPONSE rN THE P. CTNGML\S GENOME SEQUENCE...... 142
5.LI Metlrcds...... ',.,,,.,,,',,, ]42
5. 1 .2 R e sults and discus s rc,¡r? ...... ',,,,..,..,,,. 1 44
5.2 CONSTRUCTION oF A P. GING]VALIS OXYR-INACTIVATED MUTANT...... ,...... 146
5.2.1. Methods r46
5.2.1 .1 Primers and on,R amplification 146
5.2.'l .2 Construction of an o¡¡tll;;¡efQ DNA fragment '..''.'.'..',,,,,,,,,, 1 47
5.2.1 .3 P. girtgivalis transformatron...... 148
5.2.1 .4 Southern blot and PCR analysis of recombinant clones ...... 148
5.2.2 Results 149
5.3 CHARACTERIZATION OF THE P. CINGIVALISVl50 OXYR-MUTANT...... ,.,,,..152
5.3.1 Metlrcds
5.3.1.1 Effect of o-r),R disruption on anaerobic growth.... 152
5.3.1.2 Air sensitivity on solid media of OxyR and wild-type ...... 152
5.3.1.3 H2O2 sensitivity in liquid cultures...... 152
5.3.2 Results ... t 53
5.3.2.1 Effect of orl,R disruption on the anaerobic growth of P. gittgivalis ...... 153
x111 5 .3 .2.2 Air sensitivity of the OxyR mutant ...... '. I 54
5.3.2.3 H2O2 sensitivity in liquid cultures...... 155
5.4 NoRTHERN BLor ANALysrs oF rHE EXpRESSIoN oF AHpF-C tN P. ctNGIvA¿ls W50 wlLD-rYPE
AND OxYR MUTANT... 161
5.4. I M etltod,s ...... 161
5.4.2 Results
5.5 NADH OXIDASE ACTIVITY IN WILD-TYPE AND OXYR CELL EXTRACTS ....,.,.,...... 163
5. 5. 1 Methods ...... 164
5.5.2 Results t64
5.6 SUMMARY OFRESULTS AND DISCUSSION.,... 165
6. A COMPARISON OF THE TOLERANCES TO OXYGEN OF P. GINGIVALIS AND FUSOBACTERIUM NUCLEATUM AND THEIR
INTERACTIONS AS A CONTINUOUS AERATED CO.CULTURE ...... 172
6.1 COMPARISON oF THE TOLERANCES To OXYGEN OF P. GINGIVALIS AND F. NUCLEATUMWHEN
GROV/N UNDER THE SAME CONDITIONS AS CONTINUOUS, AERATED MONOCULTURES...... I74
6.1 .1 Methods t74
6.1.1.1 Microorganisms and maintenance of the strains ...... 114
6.1 .1 .2 Growth conditions 114
6. I .l .3 Gaseous atmospheres tested...... 115
6.1.1.4 NADH oxidase and SOD activity of F. nuc\eatunt...... 175
6. L 1.5 Fermentation end-products ...... 116
6. 1.2 ResuLts...... 176
6.2 THE SURVIVAL oF P, cINGIyA¿ls IN OXYCENATED ENVIRONMENTS VVHEN GROWN AS A
CONTTNUOUS AERATED CO-CULTURE WITH F. NUCLEATUM r80
6.2.1Methods 180
6.2.2 ResuLts......
6.3 CO2 REeUIREMENTS oFP. GINctvALIS AND l'. NUCLEATUM 187
6.3.1 Methods 187
6. 3.2 ResuLts and discuss io,l ...... /88
xlv 6.4 Suvvnny oF RESULTS AND DISCUSSION 191
7. GENERAL DISCUSSION 198
7.1 OxycEN ToxIcIrY To P. GINGIvA¿ls...... 199
7,2.IHE ANTI-OXIDANT DEFENCES OF P. G IN G IVA¿IS ...,..,.,,.,... 202
7 .3 F. Nucrø,qruM PROTECTS P. ct¡,tctvltls FROM oxIDATIVE DAMAGE 206
BIBLIOGRApHy ...... 21L
xv 1. Introduction l.L Periodontal diseases
Periodontal diseases comprise diverse inflammatory conditions of the tissues that support the teeth. Gingivitis is an inflammatory condition of the soft tissues surrounding the teeth while periodontitis, the most significant of these diseases, involves the destruction of the supporting structures, including not only the soft tissues but also the periodontal ligament and bone. Progressive destruction eventually leads to tooth loss (Kinane, 2001).
Periodontal diseases are a major health problem in all societies.
Periodontitis is estimated to affect at least 30Vo of the adult population, with severe forms affecting 5-6Vo (Oliver et a\.,1998). In 1989, the yearly economic burden of treating periodontitis in the USA was estimated at US$6 billion (Oliver et al.,
198e).
The pathogenic processes of periodontal diseases are the result of the host immune response to microbially-induced tissue destruction (Kinane, 2001). Of the
300 to 400 bacterial species found in human subgingival plaque samples, only 10 ro 20 species appear to play a role in the pathogenesis of periodontal diseases
(Socransky and Haffajee, 1994). The development of gingivitis is accompanied by a substantial increase in Gram-negative organisms, mainly Fusobacteriunt nucleatum, Prevotella intermedia and numerous spirochaetes (Slots and Chen,
1999).In contrast, the subgingival microflora in periodontitis, is characterised by increased proportions of Bacteroides forsythus, Porphyromonas gingivalis, and species of Prevotella, Fusobacterium, Campylobacter and Treponema (Ximenez-
Fyvie et a\.,2000).
2 Of these micloorganisms, P. gingivalis has been the main focus of research in recent years, These research efforts have provided sufficient evidence to support the existence of a strong link between this organism and the etiology of periodontitis. Apart from demonstrating that P. girtgivalis is a frequent and often major component of the subgingival plaque in patients with periodontitis, it has been shown that the microorganism is a major target of the immune response in such individuals (Kinane et al., 1999). As early as 1981, and based on antibody levels, P. gingivalis was classified as a causative agent in human periodontitis
(Mouton et a1.,19S1). This study showed that IgG antibody levels fo P. gingivalis were five times higher in adult patients with periodontitis and eight times higher in patients with generalized early-onset periodontitis compared to controls. Further studies, which demonstrated a positive correlation between IgG antibody levels to
P. gittgivalis and the severity of periodontal destruction, extended these findings and confirmed the association between P. gingivalls and periodontitis (Gmür et al.,
1986;Mooney and Kinane, 1994).
1.2 Porphyromonas gingivalis
1.2.1 General considerations
P. gingivalls is a Gram-negative, non-motile short rod or coccobacillus
(Shah and Collins, 1988). When grown on a blood agar plate, colonies appear initially as white to cream coloured, but become black after 4-8 days due to the accumulation of a porphyrin pigment on the cell surface (Holt et al., 1999). Main characteristics that define the organism are its requirement for iron in the form of
--) haem or haemin, an inability to ferment carbohydrates and an obligatory anaerobic respiration (Lamont and Jenkinson, 1998). Virulence factors of the microorganism include capsule, outer membrane vesicles, adhesins, lipopolysaccharides, haemolysins and proteinases, all thought to contribute to its pathogenicity (Holt er al.,1999).
The polysaccharide capsular material is present in most strains of P. girtgivalis, either as an amorfous or a fibrous layer (Mayrand and Holt, 1988). Its composition seems to vary among strains. However, Schifferle et aI. (1989) reported that it usually contains a high proportion of amino sugars along with glucose andlor galactose. Indeed, Farquharson et aI. (2000) characterised the capsular polysaccharide of strain ATCC 53918 finding it had peculiar, gel-like, viscoelastic properties and was comprised of mannuronic acid, glucuronic acid, galacturonic acid, galactose, and 2-acetamido-2-deoxy-D-glucose in relative molar ratios of 0.6:0.9:0.5:0.5:1.0, respectively. The presence of capsular polysaccharide material has been shown to increase the virulence of the microorganism when tested in a mouse model (Laine and Vanwinkelhoff, 1998). The hydrophilic and anionic properties of the polysaccharide contribute to decrease both the adherence to neutrophils and the deposition of complement components (Schifferle et al.,
1989). Six different capsular serotypes have been identified in P. gingivalls strains based on the presence of K antigens, probably corresponding to different capsular structures (Laine et al.,I99l). However, a link between these six serotypes and virulence has not been established.
4 1.2.2 Nutritional requirements and cytotoxic end-product formation
P. gingivalis is an asaccharolytic organism dependent on nitrogenous substrates for energy (Shah and Gharbia, 1989). Most researchers have suggested that P. girtgivalis can only utilise peptides efficiently for growth (e.g. Gharbia and
Shah 1991). However, a recent report has characterised an uptake system for the amino acids serine and threonine that indicates P. gingivalis is also able to assimilate free amino acids from the environment (Dashper et a1.,2001). Masuda et al. (2002) proposed that P. gingivalis is able to release arginine from peptides, in the extracellular environment, and metabolise it via the arginine deiminase pathway (Masuda et a\.,2001). However, no uptake system for this amino acid has yet been demonstrated.
Takahashi et al. (2000) observed that P. gingivalis consumed high proportions of the dipeptides glutamylglutamate and aspartylaspartate and determined the metabolic pathways for cytotoxic end-product formation from them. Glutamate appears to be fermented via 2-oxoglutarate and 4- hydroxybutyrate to generate butyrate or via (R)-methytmalonyl-CoA to produce propionate. In contrast, fermentation of aspaftate leads to the production of ammonia or acetate, or alternatively succinate. The latter can be converted subsequently to succinyl-CoA and incorporated into the pathways described before 'When for glutamate. P. girtgivalls was grown in Tryptone medium the main fermentation end-products were also ammonia, butyrate, acetate and propionate
Additionally, small amounts of isobutyrate and isovalerate were detected. Due to
their low molecular weights, these end-products can penetrate the periodontal
tissues and potentially disturb host cell activity and defence systems (Kurita-Ochiai
5 et al., 2000). For instance, butyrate is a potent inhibitor of the proliferation of gingival fibroblasts, potentially reducing the "wound healing" or reparative ability of these cells (B artold et al. , 1 99 1 ).
Another important feature of the growth of P. gingivalis is its preference to obtain iron from protoporphyrin D( (haem or haemin) (McKee et a1.,1986). Haem is considered to be obtained in vivo by P. gingivalis via the proteolysis of
haemoglobin and the haem-carrying plasma proteins haptoglobin and haemopexin
(Carlsson et al., 1984) and is transported into the cell in an energy-dependent
process regulated by the levels of available haemin (Genco and Dixon, 2001).
When excess haemin is available in the growth medium, haemin is accumulated on
the cell surface in the p-oxo dimeric form, [Fe(m)PPD()zO, a mechanism thought
to provide anti-oxidant protection (Smalley et al. 1998; Smalley et al. 2000).
Additionally, this ability of P. gingivalis to store haemin could provide it with an
advantage for survival in haemin-limited environments, as the accumulated haemin
can support the growth of the bacterium for at least eight generations (Rizza et al.,
1968).
1.2.3 Proteinases
P. gingivalls produces very active proteinases, whose main role is to
provide peptides for growth, although they might also contribute to invasion and
destruction of the periodontal tissues and to the modulation and evasion of the
immune system (Lamont and Jenkinson, 1998). P. gingivalis main proteinases
belong to the cysteine proteinase family, with specificity for arginine and lysine
residues, and are commonly designated Arg-gingipain and Lys-gingipain.
6 Although these enzymes have been isolated from different cell fractions, the majority of the activity is localised on the bacterial cell surface or associated with outer membrane vesicles (Holt et a|.,1999).
Two different genes have been found to encode for proteins with Arg-
gingipain activity and have been designated rgpA and rgpB. The product of rgpA is
a polyprotein that is proteolytically cleaved and separated into different domains,
an N-terminal Arg-Xaa-specific cysteine proteinase, RgpA45, and four C-terminal
adhesin domains, RgpA44, RgpA27, RgpAl7 and RgpA15 (Bhogal et al., I99l).
RgpB, the second Arg-Xaa-specific cysteine proteinase, has been isolated as a 50
kDa protein from the culture supernatant and as a 70-80 kDa membrane-associated
isoform (Slakeski et al., 1998, Rangarajan et al., 1997). This proteinase is not
associated with adhesins and does not appear to be proteolytically processed as is
RgpA (Veith et al.,2OO2).
In contrast, the Lys-gingipain activity is the product of a single gene,
designated kgp, that encodes for a polyprotein with a similar structure to RgpA,
containing the catalytic N-terminal domain Kgp48 and five adhesin associated
domains Kgp39, Kgp15, KgpI4, Kgp13 and Kgp20 (Veith et al., 2002). After
proteolytic processing of RgpA and Kgp into their individual domains, it has been
proposed that the domains aggregate via the presence of an adhesin-binding motif
to form a non-covalently associated cell-surface complex, designated the RgpA-
Kgp complex (Bhogal et a1.,1991).
Several functions have been assigned to these proteinases that appear to
contribute to the virulence of P. gingiualls. Spontaneous P. gingivalis mutants with
reduced Arg- and Lys-gingipain activity and wild-type cells treated with a
proteinase inhibitor have been reported to have a decreased virulence in animal
1 models (Kesavalu et al., 1996). Moreover, Otsrien-Simpson et al. (2001) demonstrated, using isogenic mutants in a murine lesion model, that all three proteinase genes, rgpA, rgpB and kgp, conftrbuted to the virulence of P. gittgivalis, in the following ofder, kgp >> rgpB >l= rgpA.Also, a functional kgp gene appears to be necessary for P. gingivalis binding to haemoglobin and for the accumulation of haemin on the surface of the microorganism, and in part, for haemagglutination activity (Okamoto et a\.,1998).
In vivo, the adhesin domains of the proteinases might play an important role binding to a range of host extracellular matrix proteins. Various investigators have demonstrated in vitro binding of the RgpA-Kgp complex to fibrinogen, fibronectin, laminin, haemoglobin and type V collagen (e.g. Otsrien-Simpson et al. 2001).
Binding to host proteins, might facilitate the subsequent degradation of these proteins by the catalytic domain of the proteinases (Pike er al., 1996), providing P. gingivalis with the ability to destroy tissue directly.
The proteinases are also able to modify components of the immune system.
This ability might altel the immune response to P. gingivalis or possibly constitutes a mechanism of evasion displayed by the microorganism. For example, it has been postulated that Arg-gingipain is capable of degrading complement C3 and interfering with the phagocytic events elicited by polymorphonuclear leukocytes through an inhibitory effect on the generation of active oxygen species from the activated cells (Kadowaki et al., 1994). Moreover, evidence exists that P. gingivalis proteinase activity is responsible for the degradation of human
immunoglobulins. Grenier et al. (1989) demonstrated that P. gingivalis was
capable of degrading IgG and IgA into smaller fragments, which strongly
stimulated bacterial growth in vitro.
8 Degradation of components of the immune system might also provide P. gingivalis with the ability to perturb cytokine networks itt vivo by eliminating cytokines from the local environment. Contact of epithelial and endothelial cells with P. gingivalis upregulates the mRNA of a series of mediators such as IL-IB,
tumor necrosis factor-o, IL-6, IL-8 and monocyte chemotactic protein 1 (Sandros
et aL,2000; Nassar et al., 2002). However, although up-regulated, the cytokines
are not accumulated in the cell environment, a finding that suggests they are
destroyed, after secretion, by P. gingivalls proteolytic enzymes (Zhang et al.,
ßgg).Indeed, biofilms of P. gingivalis have been demonstrated to degrade tr--lp,
IL-6 and IL-l receptor antagonist (Il--lra) (Fletcher et al., 1998). Moreover, Arg-
gingipain proteinases are able to cleave CDI4, a bacterial pattern recognition
receptor on human gingival fibroblasts, subsequently reducing lipopolysaccharide-
induced IL-8 production by these cells (Tada et al.,2OO2).
RgpB and RgpA have also been shown to efficiently activate and açgregate
platelets, suggesting a connection between the pathophysiology of periodontal and
cardiovascular diseases (Lourbakos et al., 2001).
In summary, P. girtgivalis cysteine proteinases are probably one of the main
virulence factors associated with this microorganism, as they seem to have
numerous roles such as nutrient acquisition, releasing of haem from host sources
and evasion of immune system components. In vitro results indicate that these
proteinases could also contribute directly to tissue destruction in vivo (Lamont and
Jenkinson, 1998).
9 I.2.4 Interactions of P. gingivalis with epithelial cells and other microorganisms
For P. gingivalis to survive in the oral environment, the microorganlsm needs to adhere to and interact with surfaces remote to the periodontal sulcus, including epithelial cells and other microorganisms in dental plaque.
Invasion of epithelial cells is thought to provide a protected environment that favors the colonization of this pathogen by avoiding the antimicrobial capacity of saliva and the oxygen present in extracellular sites. Indeed, P. gingivalls is able to attach and invade primary cultures of human gingival epithelial cells (Lamont et al., 1995), a process that might be facilitated by the microorganism's fimbriae
(\ù/einberg et al., 1991). Moreover, in vitro studies have shown that P. gingivalis is not only able to invade but also to multiply and persist for several days within KB epithelial cell lines (Madianos et aL.,1996).
Co-aggregation, the recognition between genetically distinct cells in suspension, and co-adhesion, the recognition between a suspended cell type and one already attached to a substratum, might be a requirement for microolganisms to colonise the oral surfaces (Kolenbrander et al., 1993). P. gingivalis has been shown to co-aggregate with a few members of dental plaque. A strong association exists between P. gingivalls and Fusobacterium nucleatum, another Gram-negative obligatory anaerobic microorganism. Co-agreggation has been demonstrated also to occur with Treponema denticola, Bacteroides forsythus (currently Tannerella forsythensis), some Prevotella species and some streptococci. The studies that have lead to the recognition of this pattern include those by Bradshaw et al., (1998), who tested the co-aggregating ability of P. gingivalis with nine other species and
10 found that the strongest co-aggregation occurred with F. nucleaturø, followed by 'Weak that with Lactobacillus rhamnos¡.¿s. interactions with Prevotella nigrescens,
Streptococcus mutants and Neisseria subflava also occurred. In another study,
Kolenbrander et al. (1989) also found that F. nucleatunt was a strong co- aggregating partner for P. gingivalis. This attachment between P. gingivalis and F. nucleatum has been characterised and shown to be modulated by a galactose- binding adhesin (Weiss et al., 2000). Co-aggregation of P. gingivalls and late colonizers has been demonstrated to occur with Treponema denticola and
Bacteroides forsythus (Kamaguchi et a|.,2001) and with Prevotella intermedia
(Kamaguchi et a\.,2001). The interaction with the latter appears to be mediated by
P. girt givalis proteinases.
P. girtgivalis benefits as well from the presence of early colonisers such as
Streptococcus gordonii. P. gingivalis is able to adhere to these microorganisms, when they have already adhered to a surface, and this co-adhesion event leads to the development of P. gingivalis biofilms. Binding of these organisms involves both the P. gingivalis major fimbrial FimA protein and the species-specific interaction of the minor fimblial Mfal protein with the streptococcal SspB protein
(Lamont et aL.,2002).
It seems that in the events that follow microbial succession in dental plaque, while physical interactions between microorganisms are important, so too is metabolic cooperation that promotes the growth of some species and modulates the immune response. Gharbia et aL, (1989) demonstrated that a growth related synergistic relationship occurred ùt vitro between P. gingivalis and F. nucleatum.
This synergistic interaction is mediated by P. gingivalis proteinases that procure peptides to F'. nucleatum, a microorganism that lacks proteinase activity but
lt possesses aminopeptidase systems (Rogers et al., 1998). Moreovet, in the murlne abscess model, Ebersole et aI., (1997) demonstrated that P. gingivalis and F. nucleatum had a synergistic effect as a mixed infection, such that their pathogenic
potential as a bacterial mixture was greater than that of either of them alone.
In summary, it appears that P. gingivalls is capable of interacting with
epithelial cells and other microorganisms, a process that might facilitate its
survival in the oral environment. Of these, the interactions with F. nucleatum
appear to be very strong at physical, metabolic and immunological levels.
1.3 Regulation of gene expression in P. gingivalis
The physiology, biological properties and pathogenicity of a
microorganism are largely influenced by its environment. Microorganisms possess
signal transduction systems that respond to specific environmental cues and ensure
that pathogens express genes required for survival (DiRita and Mekalanos, 1989).
During the transition from health to disease a number of environmental
changes are apparent in the oral ecological niches occupied by oral bacteria.
Changes in nutrients available, pH, temperature, oxygen concentration, oxidation
reduction potential (En) are likely to occur and to limit the growth of several
microorganisms (Bowden and Hamilton, 199S). An overview of how P. gingivalis
regulates gene expression in response to environmental cues other than oxygen
follows:
¡ Various studies have shown that P. gingivalis appears to alter the
expression of different genes in response to temperature changes. This response is
likely to occur also in the oral environment as different temperatures have been
t2 measured in the subgingival region in relation to tooth location and level of inflammation (Fedi and Killoy,1992). Amano et al. (1994) andX\e et al. (1991) found that P. gingivalis decreases the expression of fimbriae in response to a rise in temperature. Although the biological function of these changes is unknown, these authors hypothesised that fimbrillin expression may be more important at the early stages of colonisation (at sites with lower temperatures), facilitating adherence'
Moreover, analogs of the well characterised heat-shock proteins GroEL (HSP60 family) and DnaK (HSP70 family) have been identified in P. girtgivalis and are up- regulated in response to an increase in temperature. Shelburne et al. (2002) found that P. girtgivalis cells stressed in vitro by a 5"C temperature increase, show rapid rise in the mRNA levels of the molecular chaperons HtpG, DnaK and GroEL.
Interestingly, the expression of mRNA corresponding to RgpA was also up- regulated in response to increased temperature, a result that contradicts the previous findings of Percival et al. (1999), who showed that an increase in temperature resulted in a decrease in the total Arg-gingipain and Lys-gingipain activities, as well as in a decrease in the mRNA for RgpA and RgpB. The difference in these results could be attributed to different in vitro growth conditions and detection techniques. The former study was conducted in batch-culture-grown cells and the mRNA levels were detected by quantitative reverse transcription
PCR, while the latter study was carried out in continuous culture and the mRNA levels were detected by northern hybridization, and, as well, the activity of Arg- gingipain was assayed directly in cell extracts. Therefore, although a study carried out in continuous culture is more likely to represent the conditions of the oral environment than a study in batch culture, whether RgpA is up- or down-regulated
13 in an in vivo situation in response to the increased temperature of the inflamed periodontal pockets remains unclear.
. Changes in pH have also been seen to affect P. gingivalis expression of proteinases. Stable growth of P. gingivalis is possible in the pH range of 6.7 to 8.3,
as demonstrated in continuous culture, with optimal growth occurring at pH 7.4-
7.5 (McDermid et a\.,7988).In this study, the trypsin-like activity of P. girtgivalis
increased with increasing pH and was maximal at pH 8.0. Therefore, as the pH in
the gingival crevice changes from neutrality to alkaline, in the transition from
health to disease, it is likely that maximum expression of the cysteine proteinases
occurs in the periodontal pockets. This up-regulatory effect of pH overlaps,
however, with a possible down-regulatory effect of temperature in sites with active
disease, as discussed previously.
. Genetic regulation mediated by iron has also been studied in P. gingivalis.
Haemin-restriction regulates the expression of genes involved in haemin binding
and transport (for reviews see Forng et al. 2000 and Holt et al. 1999). As well,
Arg- and Lys-gingipain activities are regulated by the availability of haemin in the
growth medium. Haemin-restricted cultures have a lower overall level of trypsin-
like activity and a reduction in total culture biomass, although the activity present
in extracellular vesicles is greater than that of haemin-excess cultures (Smalley er
at., I99l). Moreover, McKee et al. (1986) demonstrated that P. gingivalis cells
grown under conditions of haemin-limitation were less virulent in a mouse model
than cells grown under haemin-excess, a finding that correlated with a decrease in
fimbriae production.
14 1.4 Oxygen toxicity to anaerobes and microbial response to oxygen
As a background to a later examination of the possible anti-oxidant systems operating in P. gingivalis, it is relevant, at this point, to discttss the toxic effect of oxygen on different bacteria and the most common responses displayed by microorganisms, mainly anaerobes, in lesponse to oxygen or its radicals.
1.4.1 Oxygen toxicity to anaerobes
Oxidative stress has been defined as a disturbance in the prooxidant- antioxidant balance in favor of prooxidants (Sies, 1985). Reactive oxygen species
(ROS) such as O2'-, HO' and HzOz are produced in aerobic environments and mediate cell damage (Park et al., 1992). It is well known that elevated oxygen tensions within bacterial cells increase the enzymatrc and non-enzymatic reduction of molecular oxygen to superoxide anions (Oz'-), which can form, by dismutation,
HzOz and 02. Superoxide can also react directly with iron-sulphur clusters (4Fe-
45) in enzymes, leading to their inactivation and increased levels of intracellular free iron, conducive to cell injury (Valentine et al., 1998). The reason for this is that the presence of intracellular free transition metal ions such as Fe2*, mediates the reaction of HzOz with Oz'- to form HO', probably one of the more toxic ROS, in what is known as the Fenton reaction (Rosen & Klebanoff,1919).
HzOz is also very harmful as it can easily penetrate membranes and diffuse through cells. It possesses the ability to form adducts (hydrogen-bonded chelate structures) with various cell constituents such as histidine, alanine, glycine, aspartic acid, succinic acid or DNA bases, which act as HuOz carriers (Schubert
15 and Wilmer, 1991). These characteristics allow H2O2 to act relatively far fi'om the site of its production, enhancing its oxidative effects.
The highly reactive hydroxyl radical (HO') is also generated itt vivo in other biologicatly important reactions. Activated neutrophils, for example, which release hypochlorite and superoxide may, in part, exert their cytotoxic effects on microorganisms by forming HO' in the following reaction (Candeias et a\.,1993):
HOCI t Oz'- + HO'+ Oz +Cl
ROS are toxic to the cells as they are highly reactive and can cleave nucleic acids and oxidise essential proteins and lipids (Brawn & Fridovich, i981:,Harley et at., l98l). Oxidative damage of DNA involves single strand breaks and base alterations which can induce mutations (Sigler et aI., 1999). On the other hand, protein oxidation converts several amino acid residues into their carbonyl derivatives and induces S-S and Tyr-Tyr cross-linking (Stadtman, 1992).
Modifications or splitting of protein molecules dramatically changes protein conformation, inactivates enzymes and often causes their denaturation (Sigler et at., 1999). Peroxidation of lipids generates secondary oxygen radicals such as hydroperoxides, alkoxyl and peroxyl radicals, epoxides and aldehydes, which multiply the toxic effect of the primary stress source. These secondary reactive oxygen species mainly affect proteins in the proximity of the lipids from which rhey were generated (Sigler et al., 1999). The significance of lipid peroxidation in bacteria, however, is uncertain, since most bacteria lack the polyunsaturated fatty acirls that are easily oxidised (Imlay. 2002). Still, some researchers have occasionally speculated that the debilitating effects of oxidants upon transport or energy processes are due to membrane damage. Gonzalez-Flecha and Demple,
16 (1997), for example, reported the accumulation of TBARs (thiobarbituric acid- reactive species), which are regarded as evidence of peroxidation, in E. coli strains that are deficient in alkyl hydroperoxide reductase or catalase and that, therefore, accumulate high levels of HzOz.
Aerobically growing cells are continuously attacked not only by the oxygen present in the environment, but also by internally produced ROS induced by aerobic respiration. Unlike anaerobes, aerobic bacteria are equipped with an efficient enzymatic machinery (e.g. superoxide dismutases, catalases) to decrease the level of ROS and to repair oxidative damage to biomolecules (Storz et al.,
1990). Anaerobes are incapable of surviving in oxygenated environments, as they lack the levels of these anti-oxidant enzymes needed to detoxify ROS generated after exposure to oxygen, via the mechanisms described above.
Moreover, the metabolism of anaerobes relies on metabolic schemes built around enzymes that react easily with oxygen. For example, the dependence upon low- potential flavoproteins for anaerobic respiration probably causes substantial superoxide and hydrogen peroxide to be produced when anaerobes are exposed to air (Imlay,2002).
Examples of how oxygen can alter or even eliminate certain metabolic pathways regulated by oxygen-sensitive enzymes are seen in many microorganisms. For example, in Streptococcus mlttans, the activity of the enzyme pyruvate formate lyase, involved in the catabolism of sugars, is affected by oxygen. As a consequence, the fermentation of sugars by that organism becomes homolactic, as opposed to the heterolactic fermentation carried out under anaerobic conditions (Higuchi, 19S4). Another example is seen in Fusobacterium mortiftrum, in which the fermentation of maltose is stopped when the cells are
11 exposed to air and the disaccharide maltose-6-phosphate accumulates intracellularly. This interruption in the maltose metabolism occurs because the first step in the dissimilation of the phosphorylated disaccharide appears to be catalysed by an oxygen sensitive maltose-6-phosphate hydrolase (Robrish et al.,1994).
It has become evident that the susceptibility of anaerobes to oxygen
varies even among closely-related microorganisms; and possibly correlates with
the levels of anti-oxidant enzymes present, in particular superoxide dismutase
(SOD) (Park et al., 1992). Strict anaerobes (e.g. Prevotella melaninogenica) arc
substantially more sensitive to the presence of oxygen or HzOz than, for
example, the aerotolerant anaerobe, Bacteroides fragilis or the facultative
anaerobe, Salmonella enterica serovar typhimurium, as detected by oxidative
DNA damage or by viability assay (Takeuchi et al., 1999).Indeed the lack of
adequate levels of SOD and peroxidases could be the cause of this toxicity as it
was reported by the same authors that after exposure to 02, P. melanittogenica
generated and accumulated Oz'- and HzOz, and that a crypto-OH radical
generated through HzOz was the active species mediating DNA damage
(Takeuchi et al., 2000).
1.4.2 Ãnti-oxidant enzymes
Anaerobic bacteria, including strict anaerobes, contain at least one of the
Oz- or ROS-scavenging enzymes, such as catalases, superoxide dismutases,
thioredoxin reductases, thiol peroxidase, alkylhydroperoxide reductase, or NADH
oxidase/peroxidase (Rocha et al. 1996; Rocha and Smith 1999; Caldwell and
Marquis l999;Lynch and Kuramitsu, 1999).
l8 Superoxide dismutases are widespread in eukaryotic and prokaryotic cells and have been described in the literature over the last few decades (e.9. McCord and Fridovich 1969; Miller and Britigan 1997; Shyu and Lin 1999; B,elo et aL
2000). These enzymes, which catalyse the dismutation of superoxide radicals (O2'-
+ Oz'- + 2lH+ + Oz + HzOz), constitute a family of metalloproteins with iron (Fe-
SOD), manganese (Mn-SOD) or coppef (CuZn-SOD) in their active site (Bowden
and Hamilton, 1998). In facultative anaerobic streptococci, the highly conserved
and widely-distributed Mn-SOD (Poyart et al., 1998) has been identified as the
major mechanism for detoxification of ROS. For instance, it protects Streptococcus
gordonii, from killing by paraquat (a superoxide anion generator) and by hydrogen
peroxide, and facilitates its growth in the presence of oxygen (Jakubovics et al.,
2002). Most SODs identified to date in bacteria are intracellular proteins.
However, reports of extracellular SODs also exist, for example, the Mn-SOD in
Streptococcus pyogenes (Gerlach et a\.,1998). These SODs, located upon the cell
surface, might have an even greater role in the detoxification of ROS produced by
polymorphonuclear leukocytes and macrophages in host immune responses than
cytoplasmic SODs (Miller and Britigan,1997).
Other enzymes, such as catalase and peroxidases, metabolise hydrogen
peroxide within cells, part of it the result of SOD activity (Bowden and Hamilton,
1998). Catalases, however, which convert hydrogen peroxide to water and Oz
(2IJzOz+ 2H2O + Oz.) (Neidhardt et al., 1990), are not widely distributed among
anaerobic microorganisms, being found mainly in prokaryotic cells capable of
aerobic growth. An exception is found in the aerotolerant anaerobe Bacteroides
fragilis reported to possess catalase B (Rocha et aL.,1996).
l9 Other proteins involved in HzOz detoxification are the members of the flavoprotein disulphide reductase family, which also help in maintaining the redox balance inside cells. This family consists of three distinct enzyme classes represented by glutathione reductase, thioredoxin reductase and NADH peroxidase
(Williams, 1992). All members of this family utilise redox-active disulphides in catalysis. The first of these reduction systems uses glutathione as a substrate for the
HzOz-removing enzyme, glutathione peroxidase. Glutathione reductase, which complements the system, can then redox-cycle glutathione for further HzOz removal (Miller and Britigan, 1991). Similarly, thioredoxin proteins are the electron donors for thioredoxin peroxidases (Carmel-Harel and Storz, 2000). The importance of these two systems, glutathione- and thioredoxin-based, in eukaryotic cells has been established (e.g. Meister 1992).In prokaryotes, however, although homologues of the proteins comprising the systems have been identified in E. coli and other microorganisms, including anaerobes (Harms et aL.,1998), their roles are still somewhat unclear. However, evidence exists that transcription of the genes for the above proteins in E. coli is enhanced by oxygen. As well, the two reduction systems described might modulate the activities of transcriptional regulators of the oxidative stress response (Carmel-Harel and Storz, 2000). The third member of this family, NADH peroxidase, has been well studied in Enterococcus faecalis. This protein metabolises HzOz through the oxidation of NADH (NADH + H+ + IHzOz+
NAD+ + 2HzO) and uses a non-flavin redox center, a stabilised cystein-sulphenic acid (Cys-SOH) that cycles catalytically between oxidised and reduced states, for its peroxidase activity (Poole and Claiborne, 1986).
Other enzymes involved in oxygen detoxification and in maintaining the redox balance inside cells are NADH oxidases. These enzymes convert oxygen
20 into either watel or hydrogen peroxide (NADH + H* + Oz -l NAD* * HzOz or
2NADH + 2H+ +2O2+ 2NAD* + 2H2O) (Neidhardt et a\.,1990). NADH oxidases are common in a wide range of microorgansims. They have been reported in obligatory anaerobic bacteria (Abdollahi and Wimpenny, 1990; Stanton and
Jensen, 1993), archaebacteria (Gomes and Texeira, 1998), lactic acid bacteria (Yi
et a1.,1998; Lopez de Felipe and Hugenholtz, 1999), mycoplasma (Pollack et al.,
Iggl), microaerophilic bacteria (Smith and Edwards, 1997) and protozoa lacking
mitochondria (Brown et aL.,1996). The broad distribution of this activity suggests
that NADH oxidase is a common adaptation by which microorganisms lacking a
cytochrome-mediated reduction of oxygen are able to contend with, or take
advantage of, oxygen in their environments (Stanton and Jensen, 1993). Those
NADH oxidases that have been characterised are flavoproteins, either soluble or
membrane associated. Although they seem to be involved in defence against high
levels of oxygen, they present a problem in that, since they are flavin-based
enzymes, oxygen radicals, primarily O2-', would be produced through univalent
electron transfer reactions (Massey, 1994).
Studies on the NADH oxidase from the Gram-negative anaerobic
spirochete, Brachyspira (Serpulina) lryodysenteriae, the etiologic agent of swine
dysentery, have helped to clarify the protective role of this enzyme in oxygenated
environments (Stanton and Jensen, 1993). This particular NADH oxidase appears
to be a single subunit, flavin adenine dinucleotide (FAD)-linked enzyme with a
molecular mass of about 48 kDa. Two roles for NADH oxidase in the physiology
of this microorganism have been proposed. The first is that NADH oxidase could
be an alternative mechanism for NADH regeneration, allowing the cells to take
advantage of oxygen in the environment and increase their ATP production by
21 converting acetyl coenzyme A to acetate instead of butyrate (Stanton, 1989).
However, in this role NADH oxidase appears redundant since S. hyodysenteriae
cells also contain an NADH-ferrodoxin oxidoreductase pathway for oxidising
NADH generated by enzymes of the Embden-Meyerhof-Parnas pathway.
Accordingly, the authors of these studies proposed that NADH oxidase could act,
additionally, as a defence mechanism against oxidative stress, enabling S.
hyodysenteriae to survive among oxygen-respiring and oxygen-carrying host
tissues. This hypothesis was proven by the construction, by insertional
mutagenesis, of an NADH oxidase deficient strain that was demonstrated to be
100- to 10,000-fold more sensitive to oxygen exposure than were the cells of the
wild-type strain (Stanton et a\.,1999).
In Streptococcus ntutanxs, NADH oxidase activity also appears to play an
important role in defence against oxygen toxicity and in the regulation of aerobic
mannitol metabolism (Higuchi, 1984: Higuchi, 1992). The induction of this
enzyme, and also of SOD, by the microorganism was shown to be dependent on
the energy sources and varied according to the aeration conditions. Initial
experiments indicated that two distinct NADH oxidases operate in S. mutans; an
HzO-forming NADH oxidase and an HzOz-forming NADH oxidase (Higuchi et al.,
1993). However, further characterization of these enzymes revealed that the HzOz-
forming NADH oxidase was in fact an NADH oxidase with alkyl hydroperoxide
reductase activity (Poole et a\.,2000).
These two S. mutatxs NADH oxidases exemplify the two types of NADH
oxidases so far described in bacteria. HzO-forming NADH oxidases catalyse the
four electron reduction of Oz to HzO and are exemplified by the enzyme from
Enterococcus faecalis (Matsumoto et al., 1996), which contains a stabilised
22 cysteine sulphenic acid (Cys-SOH) serving as a non-flavin redox center. These enzymes also share a high degree of sequence similarity with the Enterococcus faecalis NADH peroxidase (Ross and Claiborne, 1992) that belongs to the
flavoprotein disulphide reductase family and was mentioned before when HzOz-
scavenging systems were discussed. On the other hand, the HzOz-forming NADH
oxidases are proteins that belong to the peroxiredoxin oxidoreductase family. Apart
from catalysing the reduction of oxygen by NADH to form hydrogen peroxide,
they show extremely high peroxide reductase activity for hydrogen peroxide and
alkyl hydroperoxides in the presence of the small disulphide redox protein AhpC
(Nishiyama et a1.,2001). The HzOz-forming NADH oxidase has been called the
"AhpF" subunit. Typical examples of the AhpF-C system are found in the
facultative anaerobe, Amphibacillus xylanu.s (Niimura et al., 1993) and in the
obligate anaetobe, Bacteroides fragilis (Rocha and Smith, 1999).
1.4.3 Global regulation of the anti-oxidant response
Extensive studies conducted in Escherichia coli and Salmonella enterica
serovar typhimurium have shown that the production of antioxidant enzymes and
other stress-related proteins is regulated to some extent by environmental factors
(Harris, 1992). Because of this regulation, cells that are exposed to sublethal doses
of oxidative stress-generating agents enhance resistance to higher levels of
oxidative stress (Pomposiello and Demple, 2002). Two dimensional gel analysis of
E. coli and serovar typhimurium treated with HzOz showed that this oxidant
induces the synthesis of about 40 proteins (Morgan et al., 1986). Moreover, in a
genome-wide transcriptional profiling of the E. coli response to the superoxide
23 generating agent paraquat, 112 genes were found to be modulated (Pomposiello ¿/ al.,2O0I). Therefore, specific genetic loci co-regulate groups of proteins related to these responses, thus ensuring a rapid reaction to the environmental change. Some
of these regulons are responsive exclusively to different kinds of oxidative stress
and although the mechanisms of some have been elucidated, most of them remain
to be characterized.
The oxyR regulon is a redox-sensitive protein of the LysR family of DNA-
binding transcriptional modulators (Storz and Zheng, 2000) and functions in
cellular resistance to HzOz in E. coli by controlling the transcription of katG
(catalase), gorA (glutathione reductase), grxA (glutaredoxin-A), ahpF-C
(alkylhydroperoxide reductase), dps (protective DNA binding protein), oxys
(small, untranslated regulatory RNA), trxC (th\oredoxin 2), fur (repressor of ferric
ion uptake) and dsbG (disulphide chaperone-isomerase) (Zheng et aL.,2001). The
mechanism whereby cells sense HzOz and induce the oqtfr regulon appears to be a
direct oxidation of the OxyR protein leading to disulphide bond formation between
two cysteine residues C199 and C208 (Zheng et al., 1998). The oxidised or
reduced OxyR molecule then makes different kinds of DNA contacts: oxidised
OxyR recognises ATAGnt elements in four adjacent major grooves, while the
reduced OxyR recognises ATAGnt elements present in two pairs of adjacent major
grooves separated by one helical turn (Toledano et al., 1994), up- or down-
regulating, in this manner, the transcription of proteins according to the oxidised or
the reduced state of the environment. The mechanism by which the OxyR protein
is reduced is thought to occur via glutaredoxin-A. Since grxA is itself regulated by
OxyR, the OxyR response is then autoregulated (Storz andZheng, 2000).
24 New types of HzOz-regulators, different to OxyR, have been described recently. PerR was identified as the major regulator of the inducible peroxide stress response in Bacillus subtilis and is the prototype for a group of related peroxide- sensing repressors found in both Gram-positive and Gram-negative bacteria (Bsat et al., 1998; Mongkolsuk and Helmann, 2002). Similarly, Ohr is a family of proteins identified initially inXanthomonas canxpestris, the expression of which is strongly induced by organic peroxides (Mongkolstk et a\.,1998). Their induction requires a novel member of the MarR family of repressor proteins OhrR, the gene for which (ohrR) has been shown to be contained in the genomes of many microorganisms (Mongkolsuk and Helmann,2O02). The regulatory mechanisms of
OhrR are yet to be elucidated.
SoxR is a superoxide-responsive regulon that controls the transcription of the genes for MnSOD, endonuclease IV (for DNA repair), glucose-6-phosphate dehydrogenase (responsible for maintaining levels of NADPH), So (a ribosomal subunit protein) and OmpF (a membrane protein regulating the absorption of antibiotics). Therefore, in this case, the antioxidant protection forms part of a locus that has the capacity not only to respond to redox cycling agents, but also to decreased levels of cellular NAD(P)H, and to the presence of antibiotics (Harris,
1992).
1.5 Oxygen as a environmental cue in the oral environment and the oxidative stress response of P. gingivalís
In order to colonise the oral surfaces bacteria have to survive first in oral fluids, where they are exposed to oxygen and high oxidation-reduction potentials.
25 It is also likely that oral anaerobes encounter residual amounts of oxygen in both the early stages of dental plaque development and even in established periodontal pockets (Marquis, 1995). lndeed, the En values of a healthy gingival sulcus are around +75 mV (Kenny & Ash Jr., 1969), while periodontal pockets have been repolted to possess residual oxygen at one tenth the level in air-saturated water
(which is 0.021 pmol ml--r) (Mettraux et al., 1934) and Er, values ranging from
+14 mV to -l5l mV (Kenny & Ash Jr., 1969). Moteover, there is evidence in dental plaque of open channels that could serve to deliver oxygen deep into the biofilm (Massol-Deya et al., 1994).
It has been proposed that the survival of anaerobes in dental plaque is the result of interactions with other members of the dental plaque community. Recent studies have in fact shown that the growth of obligatory anaerobic periodontopathogens in a mixed population containing facultative anaerobes may allow them to grow in environments containing oxygen (Bradshaw et al., 1996).
Although this might be part of the process that facilitates the survival of anaerobes in plaque, it is likely that there will be many situations in which microorganisms are not protected by the overall oxygen metabolism of the microbial community; for example, passage through oral fluids or mechanical disruption of dental plaque.
In such situations, the survival of anaerobes will depend more upon defences against oxidative stress, specific for each species. Oxidative stress response mechanisms could also be important during the invasion of host tissues, as a defence against oxidative-mediated killing of neutrophils and macrophages and for survival in blood during bacteraemias. In any of these roles, anti-oxidant defence systems are likely to contribute to the virulence of anaerobic microorganisms.
26 However, little attention has been paid to studying the responses to oxygen of oral obligate anaerobes. The few studies carried out so far have shown that some
of these microorganisms might be capable of metabolising oxygen and possess
different degrees of aelotolerance. Studies on Treponema denticola revealed, for
example, that this microorganism, thought to be a strict anaetobe, can tolerate
oxygen (Syed et aI., 1993) and possesses a significant oxygen metabolism carried
out via NADH oxidase, NADH peroxidase and SOD (Caldwell and Marquis,
1999).
Amano et al. (1986) assayed the activity of SOD in cell extracts of oral
anaerobic bacteria, finding that it differed quite markedly among the species tested.
Strains of P rev otella melanino g enic a, P rev otella intermedia, P rev ot ella dentic ola
and Fusobacterium nucleatum presented the lowest activity (from 0 to 3 units/mg
protein), P. girtgivalis and Capnoclttophaga had moderate activities (around 10
units/mg protein), while Eikenella corrodens and Actinont),ces
actinomycetencomitans had the highest activity (from 26 to 12 units/mg protein).
In a further study, Amano and co-workers also reported that species of
Bacteroides, Prevotella and Porphyronlonas had also NADH oxidase activity; and
of the six species of black-pigmented bacteroides tested, P. girtgivalis showed the
highest activity of all (Amano et al., 1988). Moreover, in one strain of P.
gingivalis, NADH oxidase and SOD activities were shown to be induced, in a
moderate manner, by aeration. Further work with the SOD of P. girtgivalis 38I
demonstrated that extracts of anaerobically grown cells possessed an Fe-SOD,
while those from aerobically grown cells had a Mn-SOD, each containing three
isozymes that originated from the same apoprotein (Amano et al., 1990). Lynch
and Kuramitsu (1999) constructed a P. gingivalis mutant deficient in SOD activity,
27 demonstrating that SOD is essential for tolerance to atmospheric oxygen and appears to be protective against oxygen-dependent DNA damage. However, the
SOD mutant was no more affected than the wild-type strain in the presence of hydrogen peroxide or exogenous ROS; and was no more sensitive to killing by neutrophils. A construction of a sod::lacZ reporter transcriptional fusion also demonstrated that SOD expression is increased not only in the presence of oxygen, but also as a response to an increase in temperature.
The effect of HzOz on the batch culture growth and expression of selected proteins of P. gingivalis was studied by Leke et al. (1999). They found firstly, that the inhibitory effect of different concentrations of HzOz on the growth of P. gingívalis ATCC 49411 was dependent on the medium utilised for growth and secondly, that the presence of HzOz decreased the hemagglutinating and Arg- gingipain activities and had no effect on the expression of the heat shock proteins
GroEL and DnaK.
Other researchers have investigated the role of different proteins in the oxidative stress response of P. gingivalis. For example, Sztukowska et al. (2002) studied the role of rubrerythrin by constructing an Rbr- mutant. Rubrerythrin is a recently described non-haem iron protein, with a putative peroxide stress protective role, that is widely distributed among air-sensitive bacteria and archae-bacteria.
The P. gingivalis Rbr- mutant showed greater sensitivity to hydrogen peroxide and oxygen than the wild-type, indicating that rubrerythrin might play a role in protecting P. gingivalis against oxidative stress, either by direct reduction of HzOz or by sequestering iron from the intracellular environment.
Another mechanism protective against oxidative stress in P. gingivalis thaf involves the binding of haem to P. gingivalis cell surfaces, has been proposed by
28 Smalley et al. (1998). Formally, the term "haem" refers to reduced, ferrous or
Fe(tr) iron protoporphyrin D(, whereas the term "haemin" refers to the oxidised, ferric or FeGÐ form of the molecule. In aqueous solution, in the absence of proteins or reducing agents, iron protoporphyrin D( is found in its oxidised form
(haemin), which also exists in two forms- the ¡r-oxo fotm, a dimer with a Fe-O-Fe bridge, more likely to occur at high pH; and the hydroxide monomeric form. The hydroxide and p-oxo forms of haemin readily convert one to the other. As researchers usually add iron protoporphyrin D( in the form of haemin to the P. gingivalis growth medium, these are the forms in which haemin is present in most in vitro studies on P. gingivalis (Genco and Dixon, 2OOl). In vivo, it is more likely that P. gingivalis encounters iron protoporphyrin D( in the form of haem, as it is usually associated with haemoglobin and other host proteins. It has been proposed that when iron protoporphyrin D( binds to the cell surface, it does so in the dimeric p-oxo form (haemin) and these s,pecies can serve as an oxidative buffer system for the cells (Smalley et a\.,1998). More specifically, when iron protoporphyrin D( is released from host proteins, it is in the form of haem, Fe(tr)PPX, and will sequester any oxygen in the vicinity of the cell environment and bind onto the cell surface in the p-oxo form [Fe(Itr)PPD(]zO, through the overall reaction:
4Fe(II)PPD( * Oz -+ 2IFe(III)PPX]zO
Moreover, as formation of ¡r-oxo dimers results from a stepwise attack on the Oz molecule by Fe(II)PPD(, ROS would also form dimers as they are intermediates in the stepwise reaction (Smalley et al., 1998). Apart from reacting with oxygen and ROS, the p-oxo dimer layer will also serve as a barrier to distance
29 the outer membrane from the injurious effects of oxygen, thereby helping to
maintain an anaerobic microenvironment around the cells. In addition, all dimeric
or monomeric Fe(Itr)PPIX molecules, either bound to the cell surface or in
solution, have been shown to possess inherent catalase activity (Smalley et al.,
2000). Via this mechanism, the cells can inactivate any hydrogen peroxide present
in the extracellular environment, which could assist them to survive a neutrophil
attack.
However, the biotogical role of the accumulation of haemin l43Vo of cell
dry weight (Rizza et al., 1968)l on the cell surface is paradoxical, as any iron
stored either in or on the bacterial cell surface could have also the deleterious
effect of enhancing Fenton-type reactions that generate oxygen radicals. In this
respect, although the chemical mechansims by which the haemin layer provides
anti-oxidant protection have been described, the protective effect has not been
demonstrated in actively-growing P. girtgivalis ceIls, under controlled oxygenated
conditions.
Recently, the role of a gene involved in DNA repair after oxidative stress in
many microorganisms has been studied in P. girtgivalis. RecA is a gene involved in
P. gingivalls virulence in the murine mouse model (Liu and Fletcher, 2001) and,
although no direct evidence exist of the mechanism by which r¿cA modulates
virulence in this microorganism, it has been proposed that, as the gene is involved
in DNA repair (Fletcher et a1.,1997), its role might be to allow the cells to survive
DNA damage caused by oxidative stress during infection. Interestingly, it has been
shown that the expression of recA is increased under haem-limiting growth
conditions (Liu and Fletcher, 2001), a finding that could be linked to the fact that
30 haemin-limited cells will suffer increased susceptibility to oxidative stress because of the absence of the p-oxo dimer layer.
ln conclusion, although some information exists on the oxidative stress mechanisms associated with P. gingivalis, many questions remain. For example, the role of SOD in the organism's survival under oxidative stress has been studied, but the role of NADH oxidase and the putative peroxidase activities have not.
Moreover, neither the effect of oxygen on the expression of virulence factors or on the physiology of the microorganism, nor the molecular mechanisms regulating transcription under oxidative stress conditions have been investigated.
Furthermore, the few studies on P. gingivalis anti-oxidant enzymes have been conducted in batch culture, a situation in which close control over environmental factors is not accurate and is not likely to represent in vivo conditions. In this respect, continuous culture in a chemostat is considered to be the best ln vitro model for the study of oral microorganisms, as it most closely resembles natural conditions, in which bacteria grow at sub-maximal rates limited by the availability of one or more nutrients (van der Hoeven and de Jong, 1984).
The generation times of microorganisms in dental plaque probably vary between 8
and 12 hours (Gibbons, 1964), a situation that can be reproduced in the chemostat,
which also allows close control over environmental factors such as pH,
temperature, gas phase and En.
1.6 Summary and aims
Putative periodontal pathogens, such as P. gingivalis, exist in the supra-
and sub-gingival plaque of healthy subjects, although in low numbers and varying
3l propofiions (Ximenez-Fyvie et al.,20OO). Positive shifts in their populations that initiate disease might occur as a result of a change in local environmental conditions, due to (undisrupted) plaque accumulation over a period of time (Marsh and Bradshaw,1997). This microbial challenge initiates immune and inflammatory processes in the host tissues that, depending on host risk factors, may eventually progress to a periodontal lesion with the formation of a pocket (Kinane, 2001).
Local conditions inside the pocket then favour and prolong the survival and growth of periodontopathogens, most of which are regarded as obligate anaerobes.
However, before finding a suitable environment with optimum growth conditions, periodontal pathogens have to survive and adapt to the diverse environmental situations encountered in the different sites of the oral cavity. The presence of an unfavorable oxidation-reduction potential and the damaging effect of ROS, mainly a reflection of the oxygenated nature of the oral environment, represent the main challenge for obligatory anaerobic bacteria at the different stages of colonisation and disease progress. Little information is available on the anti-oxidant defence systems that might facilitate their survival.
Accordingly, the present research is an attempt to clarify this situation by studying the mechanisms that facilitate the survival of the anaerobic Gram- negative microorganism P. girtgivalis in response to oxygen stress. The effect of oxygen on the physiology and growth of P. gingivalis will be studied in a continuous culture system. An analysis of P. gingivalis cell morphology, fermentation end-products and proteinase activities in response to oxygen will be carried out in order to elucidate possible phenotypic changes occurring in the microorganism in an oxygenated environment. The same system will be also utilised to characterise its anti-oxidant enzyme activity. In particular, the role of
32 NADH oxidase in the oxygen metabolism of P. gingivalis will be investigated.
Furthermore, P. gingivalis genome database will be searched for putative transcriptional regulators of oxidative stress and an initial characterisation of the first oxidative stress regulon, so far identified in P. gingivalis, wlII be presented.
In the last section of these studies the oxygen tolerance of P. gingivalis will be compared with F. nucleatum, its close partner in dental plaque. Some possible interactions between these two microorganisms, in relation to their survival as a co-culture in aerated environments, will be reported.
-t -) 2. Characterisation of the growth of P. gingivalis
under oxygenated environments
34 The aim of the following set of experiments was to determine the tolerance to oxygen of P. gingivalis indlidual cells grown on agar plates and in continuous culture. The latter system was used to investigate the effect of oxygen on growth parameters, metabolic end-products and cell morphology. The effect of haemin concentration on the ability of P. girtgivalis to tolerate oxygen was also studied.
2.1 Effect of exposure to oxygen on the viability of individual cells
The aim of this initial simple experiment was to investigate the ability of P. gingivalis individual cells to survive in the presence of atmospheric oxygen, when grown on the surface of a blood agar plate. Initially, the hypothesis that temperature will have an effect on killing by oxygen was tested; and subsequently, the tolerance to exposure to air of the strain of P. gingivalis used in all of the experiments reported in this thesis, was evaluated.
2.1.1Methods
P. girtgivalis W50 (ATCC 53978) was grown in 20 mL of Brain Heart
Infusion medium (Oxoid) supplemented with 5 mg L-r of haemin (Sigma) and 0.5 g L-r of cysteine (Sigma), at 37"C, under an (anaerobic) atmosphere of Nz/HulCOz
(90:5:5). After approximately 16 hours of batch growth, serial ten-fold dilutions were performed in the same growth medium and 100 pL of suitable dilutions were spread onto anaerobic (reduced) blood agar plates (Medvet, Oxoid). Plates were exposed to air for a specified time and temperature (See Results, Tables 2.1 and
2.2). All dilutions and spreading of plates were performed in an anaerobic chamber' with a gaseous atmosphere of Nz:COz:H2 (90:5:5). After exposure to air, the plates
35 were incubated anaerobically, for 4 to 6 days and those plates containing from 20 to 200 colonies were used for colony counting. The number of colonies obtained on a plate exposed to air for a specific time was compared to that obtained on a control plate not exposed to air. The experiment was repeated twice with duplicate plates for each condition.
Two sets of experiments, following the protocol described above, were carried out. The aim of the first experiment was to determine the effect of the temperature at which the plates \ /ere exposed to air on the ability of the cells to survive. Plates were exposed to air for 1 and 2.5 hours at 3J"C and at room temperature (approx. zO"C).
In a second set of experiments all plates were exposed to air at 3J"C for time periods between 0 and 25 houls.
2.1.2 Results
Table 2.1 shows the effect of temperature on the ability of P. gingivalis
W50 to survive air exposure and form individual colonies.
Results indicated that exposure to air at 37"C had a more deleterious effect on the cells compared to exposure to air at room temperature; at which, surprisingly, after 2.5 hours no effect on the cells was detected. It was also observed that 2.5 hours of exposure to air, at 3J"C, killed only ISVo of the cell population. Therefore, a second set of experiments was carried out in which the resistance of individual cells to killing by atmospheric oxygen was further tested after more prolonged exposure times. In this set of experiments all plates were exposed to air at 31"C. Results are presented inTable 2.2.
36 Table 2.1. Survival of P. gingivalis W50 after exposure to air at
different temperatures.
R.TU 37"C
tho lj}xz%o" 97xz%o
2.5 h lOlx+Vo 82¡z7o
u Room temperature. b Time in hours of exposure to air.
'Results represent the percentage of colony-forming units (c.f.u.) appearing on plates exposed to air compared to the control
(unexposed); and are expressed as the mean + standard deviation of duplicate experiments with duplicate samples per experiment
(n=4).
)t Table 2.2. Survival of P. gingivalis W50 after exposure to air at37"C.
Timeu Survivalb
th 96xzVo
2.5 h 8U2%
5h 62x+Vo
10h 39¡zVo
20h 28xtVo
25h lTxz%o
u Time in hours of exposure to air. b Results represent the percentage of c.f.u. appearing on plates exposed to air compared to the control (unexposed); and are expressed as the mean + standard deviation of duplicate experiments with duplicate samples per experiment (n=4).
38 2.2Bffect of oxygen on P. gingivalís grown in continuous culture
The following experiments were aimed at characterising the growth of P gingivalis under oxygenated conditions, utilising a continuous culture system.
2.2.1Methods
2.2.1.1 Microorganism and maintenance of the strctitt.
P. gingivalis W50 was maintained short-term on anaerobic blood agar plates, incubated at3J"C in an atmosphere of 5VoH¡5Vo COz and 90VoNz'
2.2. 1.2 Continuous culture growth conditions.
P. gingivalis was grown under continuous culture conditions in BM medium (10 g L-t proteose peptone, 5 g L' tryptone, 5 g L' yeast extract,2.5 gL-l
KCl, pH 7.a) (Shah et al., lgl6) supplemented with 5 mg L-r of haemin, to achieve I haemin excess conditions, and 0.5 g L of cysteine. Growth in a 365 mL working- volume chemostat was initiated by inoculating the growth chamber with a 24 h batch culture of the microorganism grown in the same medium. After 16-24 h of batch culture growth in the chemostat vessel, the growth medium reservoir pump was turned on and the medium flow adjusted to give a dilution rate of 0.069 h-r 1t¿=
10 h), which was kept constant in all the experiments. This dilution rate was chosen in accordance to the mean generation time reported for bacteria in dental plaque (Gibbons, 1964). The temperature was maintained at 36"C and the pH controlled at'7 .4 by the automatic addition of 2 N KOH. These growth conditions were subsequently used for all the continuous culture experiments reported in this
39 thesis, for both P. gingivalls and F. nucleatum. The cultule was sparged with the appropriate gas mixture at a flow rate of 300 cc min-I. After about 7 generations, under all conditions, and based upon lack of change in the optical density
(ODsoonn,), the culture was considered to have achieved steady state. At this stage and for 3 consecutive days, an appropriate volume was removed fi'om the culture and analysed for its optical density, cell dry weight, and viability. Estimations of optical density were accomplished by the measurement of absorbance of the sample in a Spectronic 20 spectrometer (Bausch and Lomb, USA) at a wavelength of 560 nm. Culture dry weight estimation was performed by centrifugation of duplicate 10 mL culture samples (6,000 x g for 30 min af 4"C) in a refrigerated benchtop centrifuge (Centra MP4R, International Equipment Company, USA). The resultant pellets were washed twice in distilled water and resuspended in I mL distilled water. Cell suspensions were then placed in pre-weighed planchettes, dried at 105"C and reweighed using an analytical balance (Model H54AR, Mettler,
USA). The redox potential (E¡) of the culture at each gassing stage was monitored with a redox electrode (model Pt 4805-DPAS-SC-K85/I20, Mettler Toledo,
Switzerland). Culture purity was checked daily by Gram-staining.
Initially, P. gingivalls was grown with an incoming gaseous atmosphere of
Nz/COz (90:5) and then the incoming gas mixture was adjusted to contain increasing levels of oxygen, as follows: Nz/COzlOz (92:5:3), Nz/COz/Oz (89:5:6),
Nz/COzlOz Q5:5:10) and Air:COz (95:5).
2.2. 1.3 Statistical Analysis
Data are expressed as means * standard deviations. Differences between means were analysed for statistical significance by Student's / test
40 2.2.2 Results and discussion
Table 2.3 shows the effect of oxygen on the growth parameters of P. gingivalis W50. The organism was able to survive under all the different gaseous atmospheres tested, with the exception of Air:COz (95:5), under which the culture washed out. Under the conditions at which the culture survived, a decrease in cell viability was always observed when the gas phase was changed to contain oxygen.
This decrease in cell viability positively correlated with a reduction in the culture dry weight and optical density undel most conditions, although under the most oxygenated condition Nz/COzlOz (75:5:10), the reduced viability did not correspond with the dry weights and optical density, that were slightly increased compared to the previous oxygenated condition. This finding was accompanied by the observation, in daily Gram-stains of the culture, of a change in the cell shape of the microorganism when anaerobically grown cells were compared to those stressed with oxygen. P. gingivalis grown anaerobically presented a coccoid shape while cells stressed with oxygen grew as a bacillus (see later SEM analysis).
Therefore, the increased ratios of dry weights and optical density compared to cell viability, under l}Vo ox!flen, might be a reflection of cell elongation.
As expected, the culture E¡ increased as the oxygen concentlation in the gas phase was increased. Nevertheless, under the gas phase containing I07o oxygen, the cells were still able to create a reduced environment (Eh=-398 mV). It is worth noting that the E,n of the uninoculated medium was also measured under the different gaseous atmospheres tested. The E¡ of the medium under the anaerobic gas phasc was about -350 mV, but when oxygen was introduced, only positive E1'
4t values were recorded. Although no attempt was made to grow P. gingivalis de novo under those positive E¡ values, as the cultures were always initiated anaerobically, these values give an indication of the ability of the microorganism to reduce the environment.
The observed decrease in the culture viability under oxygenated conditions might indicate a diversion of amino acid derived ATP into cellular processes involving the maintainance of internal homeostasis. However, P. gingivalis was still able to achieve steady-state under most of the oxygenated conditions tested.
This finding makes the continuous culture system suitable for the study of the adaptive response of the microorganism to oxygen. Moreover, it further demonstrates the moderate air-tolerance of P. gingivalis'
42 Table 2.3. Effect of Oxygen on the continuous culture growth of P. gingivalis W50
Viable counts (Mean Dry Weight Optical
(mV)u + Density (56onm) Gas Phase ^Er, + SD of Log16 c.f.u. (Mean SD,
1 (Mean + mL ) n=3) mg ml,-l, n=3) SD, n=3)
Nz/COz (95:5) -507 9.94 + 0.t2 1.40 + 0.11 1.43 + 0.18
Nz/COzlOz (92:5:3) -431 9.41* + 0.10 0.95x+t 0.10 0.93*x+ 0.06
NzlCOz/Oz (89:5:6) -423 q?5**+o)) 0.83*x+ 0.07 0.91x*+ 0.06
Nz/COz/Oz (75:5: l0) -398 9.05xx + 0.18 0.85**+ 0.1 I 0.98xx+ 0.05
AirlCOz (95:5) Washed out
u E¡ (redox potential) is an average of 5 separate readings, differing from each other less than 5Vo, taken at steady state. x P < 0.05 for results compared to anaerobic growth.
*+ P < 0.001 for results compared to anaerobic growth
43 2.3 Effect of haemin on the continuous culture growth and oxygen
tolerance
In order to determine whether there is a difference in the amount of oxygen
P. gingivatis can tolerate under haemin limitation in comparison to excess, and to
test, in continuous culture grown cells, the putative protective effect upon oxidative
stress of the p-oxo layer, cells were grown under haemin-limited conditions in the
same continuous culture system utilised for the previous experiments (carried out
under haemin excess). The tolerance to oxygen of cells grown under both haemin
conditions was then compared.
The ability of a culture to metabolise a certain amount of oxygen from the
incoming gas mixture may depend on the cell numbers, if all other parameters are
maintained constant in the continuous culture system. Therefore, P. gingivalis
growth and response to oxygen under haemin limitation had to be analysed
utilising a concentration of haemin that would produce a similar viability to that of
haemin excess conditions, at the anaerobic baseline. Thus, similar starting cell
numbers between the two systems (limitation and excess) would allow direct
comparisons of the ability of the cultures to tolerate a certain amount of oxygen.
The concentration of haemin to be utilised for the haemin-limited experiment was
then likely to be a "border-line" value between haemin limitation and excess.
Therefore, an initial experiment was carried out in order to determine the optimum
haemin concentration to be utilised.
44 2.3.1 Determination of the optimum haemin concentration to be used
2.3.1.1 Methods
P. gingivalis V/50 was grown in the chemostat under anaerobic conditions using different concentrations of haemin in the growth medium. Growth medium and conditions were identical to those previously used, except where otherwise indicated. Growth was initiated by inoculating the chemostat with a 30 mL batch culture of P. gingivalis W5O grown anaerobically overnight with haemin in excess
(5 mg Lr;. The gas mixture used for all experiments was Nz/COz (90:5). Individual experiments were carried out utilising the following concentrations of haemin in the growth medium: 5, I,0.J,0.5,0.2 and 0.05 mg L-1, respectively' This range was chosen based on the findings of McKee et al. (1986) who showed that, in the same medium, haemin limited the growth of W50 up to a concentration of 0.5 mg
L-r. After 12-15 generations cultures were shown to have reached steady-state and growth parameters (viable counts and optical density) were determined.
2.3.1.2 Results
The results from these experiments are presented in Fig. 2.1, from which it can be seen that maximal limitation of growth of P. gingivalis occared when the concentration of haemin reached 0.5 mg L-t, a result similar to the one obtained by
McKee et al. (1986). This concentration was therefore used for the following haemin-limited experiments.
45 Fig.2,l Effect of haemin on the steady-state growth of P. gingivølis W50.
12 1.6
'10 1.4 1.2 I Viable 1 Optical *-** Viable Counts 6 0.8 Density (Logro (s60 counts 0.6 c.f.u. 4 nm) +oD ml') 0.4 2 0.2
0 0 0 0 'l 22.33.44.5 5 5 555 Haemin (mg L')
2.3.2 Effect of oxygen on the haemin-limited continuous culture growth of P. gingivalis
2.3.2.1 Merhods
Growth conditions and gaseous atmospheres utilised were exactly the same
as for the previous experiment carried out under haemin excess (Table 2.3), except
that the haemin concentration in the growth medium was reduced to 0.5 mg L-t.
After haemin had been added to the growth medium, an absorbance
spectrum was obtained in order to identify the state of the haemin species in the
solution. This spectrum revealed that the haemin in the growth medium ocurred as
both Fe(Itr) monomers (A¡osn') and dimers (A¡ssn.), but dimers constituted the
majority of the haem species.
46 2.3.2.2 Results and discussiott
Results are presented in Table 2.4 and are compared to resuits obtained under haemin excess conditions (Table 2.3).
From Table 2.4itcan be seen that there were no significant differences in growth parameters between haemin-excess and the chosen haemin-limited condition under anaerobic conditions. However, haemin-limited cultures seemed to have been more affected by oxygen than those grown under haemin-excess, as
shown by the decreased growth parameters and the observation that steady state
was not achieved at the most highly oxygenated condition tested (Nz/COzlOz;
75:5:10), when growth occurred under haemin-limitation. Interestingly, this culture
did not obey wash-out kinetics, but disappeared slowly until the optical density
was undetectable. This phenomenon could be explained by the appearance of thin
biofilms that formed over some of the chemostat inserts as the culture declined.
Moreover, after analysis of the cultures'E¡ under all gas phases, it can be seen that
cultures grown under haemin-excess were able to achieve a more reduced
environment compared to those that were haemin-limited.
It was also noticed that haemin-limited cell pellets presented a white
appearance, while pellets from the haemin-excess cultures had a brown colour -
indicative of the amount of haemin bound to the cell surface under the two
conditions. This haem-layer by itself, whatever its nature, might act as a protective
system that excludes oxygen from the cell surface and could be the explanation for
the increased tolerance to oxygen seen in the cells grown under haemin-excess
conditions. A more detailed discussion can be found in section 2.6.
4'l Table 2.4.Effect of oxygen on the haemin-limited growth of P. gingivalis.
Viable counts (Mean + SD of Dry Weight Optical
Gas Phase Er, (mVf Logro CFU ml-l, n=3) (Mean + SD, Density (56onm)
mg ml-l, n=3) (Mean + SD, n=3)
excesso lim' EXCESS lim EXCCSS lim EXCESS lim
Nz/COz (95:5) -501 -481 9.84 + 0.12 9.79 + 0.10 1.40 + 0.11 1.23 +0.t2 1.43 +0.18 1.34 +0.09
Nz/COzlOz (92:5:3) -431 -420 9.4t* 10.10 9.36x + 0.20 0.95x*+ 0.10 0.79xx+0.01 0.93xx +0.06 0.71*-*-+0.04
+0.06 +0.01 Nz/COulOz 189:5:6) -423 -385 9.25** + 0.22 9.15x* +0.10 0.83x*+ 0.07 0.75xx+0.08 0.91** o.l9**
NS.I Nz/COzlOz (75:5:10) -398 NSd 9.05*r'+ 0.18 NSO 0.85**+ 0.1 1 NSO 0.98*x +0.05
o E¡, (redox potential) is an average of 5 separate readings, differing from each other less than 5Va, taken at steady state.
o Data from haemin excess growth (5 mg L-l ) was presented intable2.3.
" lim - haemin limitation (0.5 mg L-r).
o NS = the culture did not survive.
x P < 0.05 for results compared to anaerobic growth under the same haemin concentration.
** P < 0.001 for results compared to anaerobic growth under the same haemin concentration
48 2.4 Morphological changes in P. gingivalis grown under anaerobic and oxygenated environments
2.4.1 SEM analysis
2.4.1.1 Methods
Continuous cultures of P. gingivalis grown as described in section 2.2,
under anaerobic and IOVo oxygen conditions, were sampled at steady-state and
prepared immediately for SEM analysis, as follows: 5 pL of cell culture were
placed on a micro glass cover slip and allowed to dry. Samples were then placed in
fixative solution containing 4Vo paraformaldehyde, l.25Vo glutaraldehyde and 4Vo
sucrose in PBS, followed by post-fixing in IVo OsOa and dehydration with
increasingly-concentrated ethanol solutions. Samples were coated, in the first
preparations, with gold, but as the coating was visible at very high magnification,
platinum was utilised instead. Coated samples were analysed using a Philips XL30
Field Emission Scanning Electron Microscope,
2.4.1.2 Results
Figures 2.2 and 2.3 show the change in the cell shape of P. gingivalis as a
consequence of exposure to oxygen. Anaerobically grown cells are coccoid while
cells stressed with oxygen clearly grow in a bacillary form, increasing by at least
three-fold in length. This change was consistent through all the continuous culture
experiments and was observed under both haemin-limitation and excess.
49 Figure 2.2 (a and b). SEM of P. gingivalls W50 grown under an anaerobic
atmosphere (Nz/COz, 90:5).
a b
Figure 2.3 (a and b). SEM of P. girtgivalis W50 grown under an
oxygenated atmosphere (N2/CO2IO2, 85:5: 10).
a b
50 2.4.2 Analysis of capsule production
2.4.2.1 Methods
To study whether oxygen had any effect on capsule production by P. gingivalis, samples from steady state cultures grown under anaerobic and
oxygenated conditions were stained with IOVo nigrosin and counter stained with
Maneval's stain (containing fuchsin and methanol). Samples were examined under
light microscopy.
2.4.2.2 Results
Utilising the staining procedure described above, it was possible to
visualise the P. gingivalis capsule but no changes were observed when cells grown
under oxygen were compared to those grown anaerobically.
2.5 Effect of oxygen on the formation of acidic metabolic end-
products
The effects of oxygen on the formation of acidic metabolic end-products by
P. gingivalis were compared for haemin-limited and haemin-excess cultures grown
anaerobically and stressed with oxygen.
2.5.1Methods
Cell-free culture filtrates were prepared from each gaseous condition at
steady-state (Tables 2.3 and 2.4) and stored at -2O"C until analysed. A high
51 pressure liquid chromatography system (HPLC) was used for acidic end-product analysis, following the methods of Guerrant et al. (1982). Briefly, the system consisted of an ion-exclusion column (Aminex Ion Exclusion, Bio-Rad), a
Rheodyne injector, a Waters 501 HPLC pump, a Waters Millipore R401 differential refractometer and a'Waters Millipore temperature control module. The chromatograms were obtained using a'Waters Millipore Data module 730. The volume of sample injected was 20 pL, the column temperature was controlled at
55oC and the flow rate adjusted to 0.5 mL min-r. The mobile phase used was 3.5 mM HzSO+.
A standard solution containing 10 mM each of formate, acetate, lactate, propionate, succinate, butyrate, isobutyrate, valerate and isovalerate was used to determine peak retention times and area for quantitation.
2.5.2 Results
Tables 2.5 and 2.6 show the effect of oxygen on the formation of acidic end-products by P. gingivalis under haemin-excess (5 mg L-r) and haemin-limited
(0.5 mg L-r) conditions.
It can be seen from Table 2.5 that the main fermentation end-product under haemin-excess and anaerobic growth was butyrate, followed by acetate. However, when oxygen was introduced into the system, acetate increased becoming the main product, while butyrate decreased. Slight changes, some of them statistically significant, also occurred in other fermentation end-products; for example, succinate and isobutyrate decreased under oxygen stress, while propionate increased. Changes in isovalerate were not consistent.
52 In contrast, the data in Table 2.6 show that under haeminlimited conditions the main end-product was acetate, which increased under oxygen stress, as it did in the haemin-excess, oxygen-stressed, culture. Butyrate, however, remained unchanged through all gas phases, while succinate, propionate and isovalerate slightly increased when oxygen was introduced into the system.
A difference observed in all gas phases, when haemin-limited growth was compared to haemin-excess, was the complete disappearance of isobutyrate under haemin-limitation.
53 Table 2.5. Effect of oxygen on the fermentation end-products of P. gíngivalis grown under haemin-excess conditions (5 mg L-1).
Gas Phase Dry Weightb Succinateu Acetate Propionate Isobutyrate Butyrate Isovalerate
Nz/COz 1.40 0.6r+0.2 1.25!0.2 2.19+0.2 2.t6+0.4 8.99+0.3 r94+0.9 (95:5)
Nz/COzlOz 0.95 0.12+0.1 9.15+0. lx 2.08+0.4 0.19+0.3x 4.82+0.2* 1.3910.8 (92:5:3)
Nz/COzlOz 0.83 0.23+0.3 13.21+0.5* 3.36+0.8 1.04+0.8 5.2010.8x 2.4t!0.3 (89:5:6)
Nz/COz/Oz 0.85 0.22+0.2 13.00+0.1* 3.28+0.1 1.28+0.2 5.58+0.4x 2.40+0.5 (75:5:10)
o End-products in mmol (g cells dry weight)-r. b Dry weights in mg ml--r. x P < 0.001 for results compared to anaerobic gas phase; n=4.
54 Table 2.6.Effectof oxygen on fermentation end-products of P. gingivalis grown under haemin-limited conditions (0.5 mg L-1)
Gas Phase Dry Weightb Succinate" Acetate Propionate Isobutyrate Butyrate Isovalerate
Nz/COz 1.23 0.15+0.1 1.55+0.2 1.63r0.6 0 5.10!0.2 r.t2+0.4
(95:5)
Nz/COzlOz 0.79 0.13+0.2 12.0910.8x 1.85r0.2 0 5.75+0.1 r.43t0.2
(92:5:3)
Nz/COz/Oz 0.15 0.43+0.3 13.44+0.6* 3.41+0.3x 0 5.70r0.3 2.20!0.2
(89:5:6)
r. o End-products in mmol (g cells dry weight) b Dry weights in mg ml--r. * P < 0.001 for results compared to anaerobic gas phase; n=4.
55 2.6 Summary of results and discussion
The experiments in section 2.1, measuring the effect of oxygen on the viability of single cells, showed that although P. gingivalls is an obligate anaerobe, it is capable of a certain degree of aerotolerance. Results from Table 2.2 showed that, although cell viability decreased as a function of time of exposure to air, P. gingivalis still displayed a moderate ability to survive. For example, I'7Vo of the total cell population survived 25 hours of exposure to atmospheric oxygen. These experiments, however, do not reveal the effect of oxygen on cell replication; that is, whether it is completely stopped or just slowed, but they do indicate that P.
gingivalis is able to withstand temporarily oxidative stress. Moreover, these
experiments illustrate intrapopulation diversity, a mechanism that could ensure
survival upon exposure to stress of a certain bacterial species (Booth, 2002). That
is, survivors of exposure to atmospheric oxygen might represent the heterogeneity
of the population with respect to some factor(s) that enables them to adapt and
probably includes enzymes involved in oxygen metabolism and DNA reparative
systems. This ability to survive the "selective" pressure of the environment would
be important for the persistence of P. gingivalis in the oral environment. Those
factors that enable single cells to survive might be the same as those operating in
the rest of the population, but are probably expressed at a different level in the
SUfVIVOTS,
It is also wofih noting that the values presented in Table 2.2, indtcating that
after 25 hours of exposure to air, 83Vo of the cells exposed had perished, are
representative only of those particular experimental conditions. For instance, lower
survival rates might have been observed if non-reduced blood plates had been
used.
56 The experiment reported in Table 2.1 also showed that temperature had an effect on killing by oxygen. This indicates that oxygen radicals play an important part in the toxic effect of molecular oxygen since, at higher temperatures, the increased reactivity of all molecules might increase their rate of generation.
Therefore, at lower temperatures (e'g' =20-30"C), there would be not only reduced oxygen-radical formation, but also the enzymes involved in providing P. gingivalis with a defence against oxidative stress would still be active. lndeed, typically these
enzymes have been assayed in vitro at 25"C (McCord and Fridovich, 1969;
Higuchi, 1984). In the clinical context, P. gingivalls cells could be less sensitive to
oxidative stress in situations of low temperature; for example, during the ingestion
of cold drinks or food. For the most part, however, P. gingivalis might inhabit
environments with temperatures around 35"C and oxidative damage would
therefore be enhanced.
When the ability of P. gingivalis to survive in the presence of oxygen was
tested in continuous culture, a decrease in cell viability was observed as the oxygen
concentration was increased. This result might indicate that oxygen produces a
diversion of cell energy into maintenance functions, such as reparing damaged
proteins and nucleic acids or metabolising ROS present in the cell environment.
When the cultures reached steady-state under the different oxygenated
atmospheres, the E¡ decreased to values around -400 mV, an indirect indication
that the oxygen present had been consumed. The only condition under which the
cells could not survive (under haemin excess) was when Air:COz (95:5) was
applied to the system. During wash-out, the Er, of this culture increased gradually,
indicating that its ability to reduce the environment had been exceeded by the
incoming oxygen in the gas phase.
57 Thus, these experiments demonstrate that not only are the cells able to cope with transient oxidative stress, as shown by the plate experiments (Section 2.1), but they are also able to survive a constant oxidative challenge.
When the effect of haemin, in excess or limitation, was tested on the ability
of P. gingivalis to withstand varying amounts of oxygen in the environment, it was
observed that haemin-limited cultures were less able to tolerate oxygen than those
grown under haemin-excess. This "protective" effect of haemin could be attributed
to the buffer capacity of the p-oxo layer of dimers formed over the surface of the
cells (SmaIley et al., 1998).In the present study, haemin in the growth medium
(pH 7.a) was already in the dimeric form (as shown by the absorbance spectrum)
aS no other reducing agents, other than cysteine, were added. ln a personal
communication, Dr. John Smalley indicated that he has shown in his laboratory
that cysteine is unable to reduce dimers to monomers and haemin will be present as
dimers at pH '7 .4. Therefore, the protection offered by the "dimer Iayef', in the
present study, is more likely to be due to the exclusion of oxygen from the surface
of the cell, rather than the reaction of monomers with oxygen. This contrasts with
the in vlvo situation, in which an additional mechanism would be operating; that is,
the reaction of haem, (Fe(tr)PPD() after being sequestered from haemoglobin, with
oxygen to form dimers that bind to the surface of the cell. Moreovet, in the system
presently utilised, the catalase-like activity present in the dimers (Smalley et al.,
2000) could have also provided some protection against any hydrogen peroxide
formed outside the cell membrane, as a result of the reaction of oxygen with some
components of the complex medium or with the cell surfaces.
The fact that haem is bound to the cells in the dimeric form is also
atlvantageous, as the dimers are almost incapable of forming ROS through Fenton
58 type reactions, because of the way the electrons of the iron are coupled to the oxygen in the molecule (antiferromagnetically) (Dr. John Smalley, personal communication). On the other hand, if monomers were present, generation of oxidants might occur, but not at the same levels as from free iron. Therefore, although haemin contains iron, Fenton-type reactions would be unlikely to occur and the p-oxo layer would allow the cells to tolerate higher levels of environmental oxygen.
Observations presented in this chapter showed that when the cells were growing in an oxygenated atmosphere they become elongated, growing as a bacillus or a short rod. This change in cell shape initially appeared contradictory, because increasing the total surface area would also increase exposure to environmental oxygen, if the internal cell volume is maintained constant. However, it is possible that the elongated cells represent in fact several cells, in which a complete septum has not been formed. If this is the case then the total area exposed to air compared to the total volume will be reduced, and the cells would have less surface area exposed to oxygen. Moreover, it is possible that elongated cells represent a mutant selected by the continuous culture procedures.
In the present study, the production of a capsular material by P. gingivalis, when grown anaerobically and under oxygenated conditions, was also investigated.
Although it has been reported that oxidative stress increases the capsular material in other microorganisms, such as Actinobacillus actinontycetemcomitans
(Scannapieco et al., 1981) and Pseudomonas aeruginosa (Sabra et a1.,2002), this was not observed in P. gittgivalis grown under oxygenated conditions.
The presence of oxygen was also seen to have an effect on the formation of acidic end-products. An increase in acetate and a decrease in butyrate were the
59 main changes noted when cells growing under haemin-excess were exposed to oxygen. In other microorganisms, acetate has also been seen to increase under oxygenated environments, co-incident with an increase in the eîzyme NADH oxidase (Stanton and Jensen, 1993). A similar shift could be occurring in P. gingivalis. According to Takahasbi et al. (2000), the formation of butyrate from aspartate would need to utilise more reduced electron carriers than the formation of acetate from the same amino acid. Therefore, in the presence of oxygen, the cells could be diverting their metabolism to the production of acetate, instead of butyrate, thus allowing the NADH oxidase to utilise the "extra" NADH for oxygen detoxification. The activity of this enzyme, NADH oxidase, will be discussed in
Section 4.
However, the decrease in butyrate could also be a consequence of the inactivation by oxygen of some of the enzymes belonging to the pathway responsible for its production. For example, 4-hydroxybutyryl-CoA dehydratase has been shown to lose its activity towards 4-hydroxybutyryl-CoA under aerobic conditions (Müh et al., l99l).Interestingly, the levels of butyrate in the present experiments were low under anaerobic conditions in haemin-limited cells. As this
reduction is certainly not due to the presence of oxygen in the environment, it
might also indicate that some of the enzymes in the butyrate pathway require iron
for activity or as part of their structure.
In summary, P. girtgivalis could be classified as an aerotolerant anaerobe.
That is, while it is incapable of colony formation on plates incubated aerobically, it
can tolerate, for a limited time, the presence of atmospheric oxygen and it will
grow in the presence of oxygen concentrations lower than that in air. Moreover, the
prcsence of a haemin-rich environment increases the tolerance to oxygen of the
60 microorganism, the growth of which under oxygenated environments is also accompanied by changes in fermentation end-products and cell morphology.
6l 3. Effect of oxygen on P. gingivalis cysteine
proteinases
62 The purpose of the study reported in this Chapter was to investigate the effect of oxygen on the activity of Arg- and Lys-proteinases (gingipains), major virulence factors in P. girtgiyalls. Assays for specific activities were performed on cell cultures grown anaerobically and in the presence of oxygen. Cell-associated proteinases were also analysed by 2-D gel electrophoresis of outer membrane preparations.
3.1 Effect of oxygen on the activity of Arg' and Lys'proteinases
The following experiments were carried out in order to examine the
possible changes occurring in the total activities (cell- and non-cell-associated) of
arginine-specific and lysine-specific proteinases, undet anaerobic conditions and
oxygen stress. Cells grown under haemin-limited and excess conditions were
included.
3.1.1Methods
When steady-state was achieved at each growth condition (Tables 2.3 and
2.4), 20 mL of cell culture were removed from the chemostat and centrifuged
(8000 x g,4"C for 30 min). Cell pellets were re-suspended in 5 mL of 0.1 M Tris-
HCl, 10 mM L-cysteine and 10 mM CaClz (pH 8.0) buffer. Protein content in both
whole cell suspensions and supematants was assayed using a BCA Protein Assay
Kit (Pierce, Rockford, IL, USA). Enzyme activity was measured in 1 mL (total
volume) of the same buffer containing 0.5 mM DL-BApNA (N-a-benzoyl-Dl-
arginine-p-nitroanilide, Sigma) or 0.25 mM AcLyspNA (N-cx-acetyl-lysine-p-
63 nitroanilide, Bachem) by measuring the free p-nitroanilide released from the
substrates when 10 pg of protein sample were added to the reaction mixture. The change in absorbance of this solution at 405 nm was recolded
spectrophotometrically, at 0.25 min intervals, for 5 minutes. A standard curve was
constructed by measuring the absorbance at 405 nm of different p-nitroanilide
concentrations and used to define units of enzyme activity. One unit was defined as
the amount of p-nitroanilide (in nmol) released by 1 mg of protein sample per
minute.
3.L.2 Results and discussion
Tables 3.1 and 3.2 show the effect of oxygen on the activity of P. gingivalis
arginine- and lysine-specific proteinases when the microorganism was cultured
under haemin-excess and haemin-limitation conditions, respectively.
The effect of oxygen on the activity of Arg-proteinase presents a complex
picture. There is an increase in the activity of the Arg-proteinase present in the
whole cells fraction and a decrease in the activity found in the supernatants,
irrespective of the haemin concentration. However, if a total activity is obtained by
adding the activity present in cell pellets and supernatants, there does not seem to
be an increase, except for the culture grown under 3Vo oxygen Therefore, it
appears that oxygen, rather than having an effect on the expression of the enzyme,
seems to alter its location.
On the other hand, the activity of Lys-proteinase appears to decrease in
both whole cells and supernatants as oxygen is introduced into the system. This
observation was again consistent under both haemin concentrations, although
64 under haemin-limitation the decrease in Lys-gingipain did not reach statistically significant levels
65 Table 3.1. Effect of oxygen on proteinase activity of P. gingivalis when gro\ryn
under haemin excess conditions (5 mg L'1).
Gas Phase Lys-proteinase activityu Arg-proteinase activityu
CP SN CP SN
Nz/COz 190+ 13 t32+ 20 606 + 16 259 ! tr
(95:5)
NzlCOz/Oz t16+8 85 +34 89t* + 52 98* + 29
(92:5:3)
NzlCOz/Oz 138t + 9 69*+ l 150* + 32 101x + 13
(89:5:6)
Nz/COzlOu I36x + 3 64x+8 185* + 26 83* + 44
(75:5:10)
u Specific activity associated with the re-suspended cell pellets (CP) and -t), supernatants (SN) is presented as mean + SD (nmol (mg protein)-r min n=4.
* P < 0.05 for results compared to anaerobic growth.
66 Table 3.2. Effect of oxygen on proteinase activity of P. gingivøJls when gro\iln
under haemin-limited conditions (0.5 mg L-1).
Gas Phase Lys-proteinase activityu Arg-proteinase activityu
CP SN CP SN
Nz/COz 103+ Il 20!5 406 + t5 r75 + 2r
(95:5)
Nz/COzlOz 82+7 16+8 534* + 32 76* + I'7
(92:5:3)
Nz/COzlOz 19 !6 15+3 502* + 4I 64x + 18
(89:5:6)
u Specific activity associated with the re-suspended cell pellets (CP) and
I -r), supernatants (SN) is presented as mean + SD (nmol (mg protein) min n=4. x P < 0.05 for results compared to anaelobic growth.
67 3.2 Anatysis of proteinases in P. gingivølis outer membranes by 2'D gel electrophoresis
As Arg-gingipain activity is the result of two proteins encoded by different genes, and since both Arg- and Lys-gingipain are proteolytically processed after the polyproteins are translated, a proteomic analysis of P. gingivalis ottet membrane fraction was considered necessary to confirm the findings of the
enzyme assays and to determine possible changes occurring in the processing of
the enzymes. Moreover, changes in the expression of other outer membrane
proteins, in response to the changed atmospheric conditions, were also
investigated.
3.2.1Methods
3.2.1.1 Grovvth and harvest o.f P. gingivalis.
50 mL of P. gittgivalis cells grown under haemin-excess conditions, in an
anaerobic atmosphere of Nz/COz (95:5) and under the oxygenated condition
Nz/COzlOz (75:5:10), were obtained directly from the chemostat vessel at steady
state and harvested by centrifugation (6000 x g, 4"C for 30 min) in the presence of
1 mM tosyl-L-lysine chloromethyl ketone (TLCK), an inhibitor of cysteine
preoteinase activity.
3.2. 1.2 Outer membrane preparation.
The outer membrane (OM) of P. girtgivalis was prepared by the sarcosinate
method (Filip et a1.,1913) as modified by Veith et al., (2001), Briefly, harvested
68 cells were re-suspended in 10 mL of buffer (50 mM Tris pH 8.0, 150 mM NaCl, 5 mM MgClz) and 300 units of RNase and DNase were added. Cells were then sonicated on ice for 15 min with the addition of TLCK (1 mM final concentration) at the beginning, during and at the end of sonication. Unbroken cells were removed by centrifugation (4000 x g,4"C for 15 min). The membranes were then collected after centrifugation (40000 x g, 4"C for t h) and washed twice with buffer containing lmM TLCK. Membranes were re-suspended in an equal volume of buffer and 2Vo sodium-lauryl sarcosinate and incubated at 37"C for t h to facilitate
inner membrane solubilisation. The OMs were pelleted at 40000 x g, washed with buffer containing 1 mM TLCK and solubilised in IPG solution (7 M urea, 2 M
thiourea, 4Vo wlv CHAPS. 0.5Vo Triton X-100, 50 mM DTT and 0.27o canier 3-10
ampholytes). Solubilisation was allowed to proceed for 2-3 h with intermittent
vortex mixing. Samples were then centrifuged (14000 x g for 15 min) to remove
insoluble material. Protein content was determined with a Coonassie Plus Protein
Assay Reagent Kit (Pierce, Rockford, IL) using bovine serum albumin as a
standard. Samples were diluted 1/10 to avoid interference with the protein assay.
3.2.1.3 2D PAGE
Seven cm immobilised pH gradient (IPG) strips (pH 3-10) were rehydrated
actively at 50 V for 16 h with a volume equivalent to 100 pg of protein (diluted in
IPG buffel solution). Isoelectric focusing (IEF) was caried out using a Protean IEF
Cell System (BioRad) at 20"C under paraffin oil. An IEF cycle consisted of a
conditioning step af 250 V for 15 min to remove salt ions and charged
contaminants; a voltage ramping step in linear mode; and a final focusing step, at
which strips were focused at 4,000V for at least 20,000 V/hours. IPG strips were
69 then prepared for the second dimension by a 10 min incubation in reducing equilibration buffer (6M urea, 207o SDS, 0.05M Tris/HCl pH 8.8, 2O7o glycetol)
containing 2Vo (w/v) DTT, followed by a 10 min incubation in alkylating equilibration buffer containing 2.57o (wlv) iodoacetamide. SDS/PAGE was then conducted in l2Vo polyacrylamide gels using the standard Tris/glycine buffer
system of Laemmli (1970). IPG strips were placed in direct contact with the
second dimension gel and sealed with IVo molten agarose containing bromophenol
blue. Two gels (one for each condition) were run simultaneously at 70 V until the
tracking dye passed the stacking gel and then at 125 V until the tracking dye
reached the bottom of the gel.
Gels were stained with Coomassie brilliant blue or with Sypro-Ruby stain
(BioRad) according to the manufacturer's instructions.
3.2.1.4 In-gel digestion of proteins separated by 2D PAGE
Trypsin digestion of protein spots was performed following the protocol
described by Veith et al. (2001). Coomassie blue-stained spots were cut from the
gel and washed in I mL 100 mM NH+HCO¡, pH 8.0, containing i mM CaClz and
25 ng pl--r sequencing-grade modified trypsin (Promega). After rehydration (=lQ
min), excess trypsin was removed and the gel pieces kept moist by pipetting2O ¡tL
buffer into the cap of the tube. The rehydrated gel pieces were then incubated
overnight at 3J"C. The peptides in the gel pieces were first extracted in 50 pL 20
mM NH¿HCO3, pH 8.0, followed by 3 extractions in 50 ¡tL 5Vo formic acid/ 5OVo
acetonitrile / 45Vo HzO (v/v/v). Each extraction involved sonication for 2O min. The
extracts were combined and dried in a Jouan RC10.10 rotary evaporator.
10 3.2.1.5 Peptide mass fingerprinting (PMF).
PMF was performed using a Voyager DErM matrix assisted laser-desorption ionization-time of flight (MALDI-TOF) mass spectrometer (Applied Biosystems,
Foster City, CA, USA) in linear mode with a delay time of 100 ns and an acceleration voltage of 20 kV. Dried extracts were dissolved in 307o acetonitrilellVo formic acid. Next, a 0.8 pL solution of 0.2Vo (w/v) nitrocellulose and saturated matrix in acetone was applied to the sample plate followed by 0.3 pL of peptide extract. After drying, the sample was washed twice with 2 ¡tL lVo formic acid. The matrix used was a-cyano-4-hydroxycinnamic acid and calibration was performed externally with angiotensin tr and ACTH 18-39 human clip
(Auspep, Melbourne, Australia).
3.2. 1 .6 P rotein identification
For protein identification, a database previously produced at the School of
Dental Science (The University of Melbourne) was utilised (Veith et a1.,2001).
For database construction, the preliminary DNA sequence of P. gingivalis W83
genome was obtained from the TIGR website at http:llwww.tigr.org. The DNA
sequence was then translated to protein in all six reading frames and a database of
more than 128 000 putative proteins lvas generated. Proteins were identified from
peptide mass data using the program General Protein Mass Analysis for Windows
(GPMAV/) (Hojrup).
11 3.2.2 Results and discussion
Figures 3.1 and 3.2 show two representative examples of 2D gels of the
OM of P. gingivalls grown under an anaerobic atmosphere and in the presence of
IOTo ox!fleî.
Gels were reproducible and, in general, the pattern of spots between the two conditions was comparable. Gels were also comparable to those published by
Veith et al. (2002) which were useful to identify, prior to PMF, those protein
"spots" possibly corresponding to cysteine proteinases. Other proteins, such as
Omp 40 and Omp 41, were also identified by PMF and served as reference points
when comparing gels. Protein spots differing in one of the two gels were also
included in the PMF analysis. Table 3.3 shows the PMF identification data for all
protein spots.
As will be discussed later, several protein spots gave a much lower than the
expected molecular weight. These spots corresponded to fragments of a specific
protein, of which the full length version was also identified in the same gel. These
fragments could be the result of incomplete inhibition by TLCK of the cysteine
proteinases, thus resulting in partial protein degradation during the OM
preparation. An example of this is Omp 40, two fragments of which were identified
at the low molecular weight region, while the full length version appeared at the
expected region of the gel. Nevertheless, the reproducibility of the gels between
the different batches of cells tested allowed valid comparisons to be made.
12 3.2.2.1 Results for Arg-proteinase (RgpA and RgpB)
Different protein spots were identified as RgpA, corresponding to some of the previously reported domains of the protein after proteolytic processing (Veith et al., 2002). The c-termini adhesin domain, RgpA44, migrated together with a protein identified as PZl. Similar results were obtained previously by Yeith et al.
(2002). This protein spot appeared with increased intensity in the OM gels of cells
stressed with oxygen. However, the RgpA catalytic domain, RgpA45, identified as
a train of spots in the gels of the anaerobic conditions, seemed to have disappeared
in the oxygen gels. Such results with RgpA45, however, could be an artefact
produced by the proteolytic cleavage, as there is a protein spot containing
fragments of RgpA directly below RgpA45, in a lower molecular weight region.
RgpB appears as a single spot that migrates together with a putative
lipoprotein named P59. The intensity of this spot containing RgpB is increased in
the "oxygen" gels.
Moreover, the "oxygen" gels contained two spots of low intensity and high
molecular weight that were identified as RgpA. The trypsin digest of these spots
produced peptides that matched both the catalytic and the c-termini adhesin
domains, as well as other regions of the protein. It is likely that these spots
corresponded to unprocessed RgpA, as its molecular weight closely approaches
that expected for the uncleaved polyprotein.
3.2.2.2 Results for Lys-proteinase (Kgp).
Two spots were identified as Kgp. The c-termini-adhesin domain, Kgp39,
migrated close to another protein spot containing RgpA44 andPZJ .B,ecatse of this
it was difficult to determine whether its apparent increased intensity in the
t-) "oxygen" gel was due to an artefact or whether it reflected a real increase in expression. On the other hand, the catalytic domain, Kgp48, appeared as 2 or 3
spots on a train, which showed decreased intensity in the "oxygen" gel. This result
is in accordance with the finding that oxygen stressed cells have a decreased Lys-
proteinase activity compared to cells grown anaerobically.
3.2.2.3 Other findings
Several spots migrating at the same pI and with molecular weights between
30 to 34 Kd were identified as HagA, a haemagglutinin adhesin precursor. The
HagA polyprotein has a theoretical molecular weight of =230 kDa, but suffers
post-translational proteolytic cleavage as RgpA and Kgp. The proposed structure
(Veith et a\.,2002) is as follows: HagA signal peptide (nucleotides 1-25), HagA35
(26-351), HagA3O (366-625), HagAl8 (658-803), HagA32 (820-l0ll), HagAl8
(1137 -1255), HagA32 (121 2-1559), HagA 1 8 (1589 -ll O1), Hag Al5 (11 24- 1 856),
HagAl3 (1859-1978), Hag 420 (1980-2164). Several of these domains were
identified in the OM gels and the intensity of the spots appears to increase
markedly in the oxygen gels, suggesting an increased expression of HagA under
oxygenated conditions.
14 Figures 3.1 and 3.2. 2D-gels of the OM of P. gíngivalis gro\ryn under an anaerobic atmosphere and in the presence of l07o oxygen' respectively.
Figure 3.1
pH3 pH 11
66
RgpA44 P2'1 55
Kgp39
HagA
RgpB, RagB and RgpA fragments
Figure 3.2
pH3 pH 11 Kda 200
RgpA44 P2l
Kgp39
Omp40/41
HagA
fragment RgpB. Rag B and RgpA fagments
15 Table 3.3. Identification data for protein spots in Figs. 3.1 and 3.2
Observed mass Peptides Accession
Protein spot (kDa) matched/expecteda Description rrumberb
RgpB 67 -95 t4/17 Arginine-specific cysteine proteinase: gingipain R-2 Pg046l
RgpB fragment 27-31 5il7 Arginine-specific cysteine proteinase: gingipain R-2 Pg046l
P59 61-95 t5l2l LPS-modified surface protein Pgl 838
RgpA t78ll14 It-nt23 Arginine-specific cysteine proteinase gingipain R-l Pg 1768 RgpA44 46-59 6t7 Arginine-specific cysteine proteinase gingipain R- I Pg 1768 RgpA45 43-41 I t/t4 Arginine-specifìc cysteine proteinase gingipain R- I Pg 1768 RgpA fiagrnent 27-3t 2t23 Arginine-specific cysteine proteinase gingipain R- I Pg 1768
P27 46-59 5i8 LPS-modifi ed surface protein Pg I 570
Kgp48 48 t0il4 Lysine-specific cysteine proteinase Pgl 605
Kgp39 43 6u Lysine-specific cysteine proteinase Pg I 605
HagA (several spots) 30-34 Hemaggl utinin/adhesin precursor P91602
Omp40 43 t5lt6 Putative porin Pg0626
Omp4l 43 14il6 Putative porin P90627
Omp 40 n-terrninal 30 4t16 Putative porin PsO626
domain
Omp 40 fragment 29 5/16 Putative porin Ps0626
Rag B fiagment 21-31 2^9 Outer membrane lipoprotein; receptor antigen B PgOl7 | u The number of peptides marched, including fully cleaved peptides and partially digested fragments, within a mass range of 1000-5000 Da re lative to the total number of fully cleaved peptides in the same range in the theoretical tryptic digest. b Accession numbers correspond to the numbers fiom the Oral Pathogen Sequence Databases obtained from http://www.oralgen.lanl.gov
76 3.3 Summary of results and discussion
A comparison of the Arg-proteinase activity between haemin-limited and excess cultures further confirmed that the cells grown with 0.5 mg L-l of haemin were haemin-limited, as the Arg-proteinase activity was lower than at 5 mg L-1.
This result is in accord with previous repofts on the activity of this enzyme under conditions of haemin-excess and limitation (Marsh et al., 1994).
Basically, the enzyme assays showed that the cell-associated Arg-
proteinase activity was increased by oxygen stress, while the corresponding
supernatant activity decreased. On the other hand, Lys-proteinase activity
decreased in both fractions under oxygenated environments. Unfortunately, the
data obtained from the 2-D gels, although reflecting in part the results of the
eîzyme assays, should be qualified as the proteinase activity was not completely
inhibited and might have cleaved some of the proteins during the OM preparation.
Time and budget constrains (TLCK is an expensive chemical) limited repetition of
these gels.
Neveltheless, results for Arg-proteinase (RgpA and RgpB) in the 2-D gels
show an increase in the expression of the adhesin domain of RgpA and an increase
in RgpB. This result seems consistent with the results from the assays, although the
disappearance of the catalytic domain of RgpA in the gel of oxygen-stressed cells
does not agree with these observations. Perhaps the RgpA catalytic domain was
preferentially cleaved by the proteinases remaining active during outer membrane
preparation. Alternatively, the increase in cell-associated Arg-gingipain activity
could be a result of the increase of RgpB only. However, this explanation is less
likely because an increase in the adhesin domain expression of RgpA is expected
71 to occur simultaneously with an increase in the catalytic domain, as rgpA is transcribed as a single mRNA molecule (Percival et aL.,1999).
The results for Lys-gingipain appear to be more consistent. There was a decrease in the intensity of the catalytic domain in the 2-D gels, and the assays showed also a decreased activity.
The reasons for these changes in the cysteine proteinases remains unclear.
Oxygen does not seem to have an effect on the overall Arg-gingipain activity, but more cell-associated enzyme(s) was found in the oxygen-stressed cells. Perhaps, oxygen could affect the processing of Arg-gingipain in such a way that it impedes enzyme liberation into the extracellular environment. The presence of unprocessed
RgpA spots, running in the high molecular weight region of the "oxygen" gels, could be an indication of this. One could speculate that when present in the
oxygenated oral fluids, or during dental plaque maturation, the liberation of
peptides from proteins would be more beneficial if it occured in close proximity to
the microorganism. In this respect, increasing the Arg-gingipain cell-associated
activity could replesent a nutritional advantage, as the cells are not confined to a
limited spatial niche as occurs in periodontal pockets. Maintaining the levels of
Arg-gingipain under oxidative stress could have also an additional significance in
vivo as this enzyme has been shown to be responsible for interfering with the
activity of polymorphonuclear leukocytes (Kadowaki et al., 1994), that generate
active oxygen species.
On the other hand, the decrease in the total Lys-gingipain activity could
represent a real decrease in the transcription or translation of kgp or, alternatively,
oxygen could have affected the catalytic properties of the enzyme, perhaps due to
78 protein oxidation. Unfortunately, the 2-D gels do not solve this question and the significance of such a finding is not clear.
Nevertheless, it seems that in the presence of moderate amounts of oxygen,
P. gingivalls is capable of retaining most of its proteolytic activity, an indication that the cells might still be able to display virulence in moderately oxygenated environments.
The potential for the increase in the expression of HagA is also an interesting result revealed by the 2-D gels. The haemagglutinin gene, hagA, possesses adhesin domain regions responsible for haemagglutination and haemoglobin binding that are also located in the C-terminal regions of rgpA and kgp (Shi et al., 1999).It would seem logical that if the accumulation of haem on the surface of the cells is advantageous as a defence against oxidative stress, then the cells would favour the acquisition of this molecule from haemoglobin.
However, the decrease in Lys-gingipain, also involved in haemoglobin binding, would contradict these observations. Thus, at this time, it is not clear why a potential increase in the HagA protein would be beneficial for the cells in a
situation of oxidative stress.
19 4. Studies on enzymes involved in the metabolism of oxygen by P. gingivalis
80 4.1 The effect of oxygen on the activity of anti'oxidant enzymes
The activity of the enzymes NADH oxidase, NADH peroxidase and SOD,
involved in the detoxification of oxygen or its radicals, was assayed in cells grown
in continuous culture, under anaerobic and aerated environments.
4.1.1Methods
4.1.1.1 Preparation ofcrude cell extracts.
Growth conditions in the chemostat were the same as those described in
sections 2.2.I.2 and 2.3.2.1. At steady-state, under haemin-excess and haemin-
limited anaerobic and aerated conditions (see Tables 2.3 and 2.4), the chemostat
was sampled. Bacterial cells were harvested by centrifugation (6,000 x g, 4"C, for
30 min) and washed twice with 50 mM potassium phosphate buffer, pH 7.8,
containing 0.1 mM EDTA. The cells were then re-suspended in an aliquot of the
same buffer and lysed by ten, l5-second sonications on ice. The disrupted
suspensions were centrifuged (6,000 x g, 4oC for 30 min) and the protein content of
the supernatants was determined with a Coomassie Plus Protein Assay Reagent Kit
(Pierce, Rockford, IL), using bovine serum albumin as a standard.
4.1.1.2 Enzyme Assays.
NADH oxidase and NADH peroxidase activities were assayed at 25"C
following the methods of Higuchi (1992). NADH oxidase was assayed by
monitoring, spectrophotometrically, the oxidation of B-NADH at A3a6n.. The
reaction mixture (3 mL) contained 50 mM potassium phosphate buffer (pH 7.0),
8t 0.1 mM P-NADH and cell extract (0.125 mg protein). NADH peroxidase was assayed under anaerobic conditions, achieved in a Thunberg-type cuvette, using the same reaction mixture as for NADH oxidase, but with the addition of 0.3 mM
HzOz.One unit of activity for both enzymes was defined as the amount of extract that catalysed the oxidation of 1 nmol of NADH min-|. SOD activity was measured at 550 nm by competitively inhibiting the reduction of cytochrome c at 25"C, following the methods of McCord and Fridovich (1969). The reaction mixture (3 mL) contained 50 mM potassium phosphate buffer (pH 7.8), 0.1 mM EDTA, 0.01 mM cytochrome c,0.1 mM xanthine, cell extract (0.125 mg protein) and sufficient volume of xanthine oxidase to produce a constant reduction of cytochrome c af a rate of 0.02 of absorbance increase min-I. One unit of SOD was defined as the amount of extract that decreased by 5O7o the rate of reduction of cytochrome c.
4.1.1.3 Further analysis of NADH oxidase activity.
The effect of flavin adenine dinucleotide (FAD) as a possible co-factor of
NADH oxidase was analysed by adding 0.02 mM FAD to the NADH oxidase reaction mixture.
To determine if NADPH could be used as an alternate substrate for the
oxidase, 0.1 mM of p-NADPH was added to the reaction mixture instead of
NADH.
According to Caldwell and Marquis (1999), NADH peroxidase activity can
be suppressed by the addition of excess catalase to reaction mixtures to allow for
the determination of the oxidase activity only (in case the NADH oxidase present
yields HzOù. Therefore, catalase was added to the NADH oxidase reaction mixture
82 to a final concentration of 2300 units ml-land this reaction was compared to one without added catalase
4.1.2 Results and discussion
4.1.2.1 Effect of O2 on the activity of anti-oxidant enzymes of P. gingivalis grown
under haemin- exc e s s and haemin-Iimited c onditions.
The activity of anti-oxidant enzymes detected in P. gingivalis is presented in Tables 4.1 and 4.2. Activity of each of the three enzymes tested (NADH oxidase, NADH peroxidase and SOD) was detected under all conditions and mostly the activities increased under oxygenated environments compared to those obtained for anaerobically grown cells. NADH oxidase activity increased with oxygen in both haemin-limited and haemin-excess cultures. However, the magnitude of increase in activity, after a determined oxygen concentration was applied to the culture, was higher under haemin-limitation. For instance, the increase in activity under haemin-limitation from anaerobic to 67o oxygen was 1.4 fold, while under haemin-excess the activity of the eîzyme increased only 0.8 fold.
NADH peroxidase activity behaved similarly, doubling from anaerobic growth to
6Vo ox!flen under haemin-limitation, while the increase under haemin-excess was only 0.4 fold. Thus, if the increase in the activity of these enzymes reflects the environmental pressure exerted on the cells, these results seem to confirm that the
cultures grown under haemin-limitation were more stressed by oxygen than those
grown under haemin-excess. The haem layer formed over the surface of haemin-
excess cells might isolate the cells from exposure to oxygen as well as functioning
'Whereas, as a peroxidase-like system (Smalley et a\.,2000). haemin-limited cells
83 probably increased the expression of anti-oxidant enzymes to compensate for the lack of the haem layer.
On the other hand, SOD activity increased only slightly under the oxygenated atmospheres tested. In this respect, these results differ with those obtained previously by Amano et al. (1988) who reported that, in batch gro\ /n cells, the SOD activity of P. gingivalis 381 was induced more by aeration than was its NADH oxidase activity.
Table 4.1. Effect of Oz on the activity of anti-oxidant enzymes of P. gingivalis
under haemin-excess (5 mg L-1) conditions.
Gas Eh NADH NADH
Condition (mv) oxidase peroxidase SODb
Nz/COz -507 5.49 + 0.51" 8.41 + 0.61 8.94 + 0.05
(95:5)
Nz/COzlOz -431 7.08 + 1.11 8.33 + 0.51 10.44 + 0.51*
(92:5:3)
Nz/COzlOz -423 10.3 + I.23** 12.53 + 0.81** 10.71 + 0.33*
(89:5:6)
NzlCOzlOz -398 9.18 ! 1.95** 12.29 * Q.lQx* 10.83 + 0.29**
(75:5:10)
uMean t SD of activity (Units (mg protein)- ), n=4
b Superoxide dismutase.
-P < 0.05 for results of anaerobic versus 3Vo,6Vo or lOVo Oz.
**P < 0.001 for results of anaerobic versus 3Vo,6Vo or lOVo 02
84 Table 4.2.E,ffect of Ozon the activity of anti-oxidant enzymes of P. gingivalis
under haemin-limited (0.5 mg L-1) conditions.
Gas Phase Eh NADH NADH SODb
(mv) oxidase peroxidase
Nz/COz -487 g.r5+0.12" 9.04+0.81 8.55+0.62
(95:5)
Nz/COzlOz -420 12.77+r.43" rr.r2+t.06.. 8.45+0.52
(92:5:3)
Nz/COzlOz -385 19.83t1.18"" 18.11+1.82.. 10.67+0.57.
(89:5:6)
oMean + SD of activity (Units (mg protein)- ), n=4
b Superoxide dismutase.
-P < 0.05 for results of anaerobic versus 3Vo 02 or 6VoOz
--P < 0.001 for results of anaerobic versus 3Vo 02 or 6VoO2.
4.1.2.2 Further analysis of NADH oxidase activity
4.I.2.2.1 Flavin requirement.
As shown in Tables 4.1 and 4.2, extracts of P. gingivalis could catalyse the aerobic reduction of NADH without the requirement of added flavin. However, as seen in Figure 4.1, addition of 0.2 mM FAD enhanced the NADH oxidase activity of cell extracts. This dependence on added flavin was more marked when cell extracts were frozen and thawed repeatedly. Freezing and thawing of cell extracts
85 resulted in a decrease in activity that could only be restored by added FAD, which indicates that FAD serves as a co-factor in the NADH oxidase reaction and that the freezing and thawing process probably damages or produces a dissociation of the co-factor that inactivates the enzyme. These observations are similar to those obtained by Caldwell and Marquis (1999) who tested, in a similar fashion, the
NADH oxidase activity of Treponema denticola cell extracts.
Figure 4.1. NADH oxidase (Nox) activity in the absence and presence of 0.2
mM FAD
0.52 0.5 0.48 *Nox A34on,n O.Æ +Nox+ 0.4 FAD 0.42 0.4 012 3 4 5 6 7I Time (min)
86 4.1.2.2.2 NADPH as a possible substrate for oxidase activity
As shown in Figure 4.2, NADPH was less efficient than NADH as the source of reducing equivalents for oxidase activity.
Figure 4.2. Oxidase activity utilising NADPH vs. NADH as a substrate
0.52 0.51 +NADH 0.5 oxidation 0.49 A34o n.n --**NADPH 0.48 oxidation 0.47 0.46 0.45 02468
Time (min)
4.1.2.2.3 NADH oxidase and NADH peroxidase activities
As mentioned earlier, the following logic has been utilised previously by
Caldwell and Marquis (1999) to determine whether the end-product of NADH oxidase is HzO or HzOz. If one assumes that NADH oxidase and NADH peroxidase are independent systems, any H2O2 produced by an H2O2-yielding
NADH oxidase, will serve as a substrate for NADH peroxidase. Therefore, addition of catalase to the NADH oxidase reaction allows assay exclusively for the
NADH oxidase activity in cell extracts since catalase will immediately metabolise any HzOz produced by NADH oxidase. Consequently, NADH peroxidase will have no available substrate. Thus, in the presence of an HzOz-
87 yielding NADH oxidase, a decrease in the rate of oxidation of NADH is expected when catalase is added to the reaction mixture.
Figure 4.3 shows the effect of adding catalase to the P. gingivalis NADH oxidase reaction. It can be seen that extracts oxidised NADH at the same, or a slightly faster, rate when catalase was added to the reaction mixture. This result suggested that significant amounts of HzOz were not produced by the NADH oxidase reaction and so the product of NADH oxidase activity was therefore more likely to be HzO. These results will be discussed later (Section 4.6).
Figure 4.3. NADH oxidase (Nox) assayed aerobically in the presence of
2300 units of catalase ml,'l.
0.52
0.51
0.5 A34o nnt 049 *Nox 0.48 *#* Nox+Catalase o.47
0.46
0.45 0 2 4 6 I Time (min)
88 4.2lsolation and characterisation of NADH oxidase activity
To identify the protein(s) responsible for the NADH oxidase activity in P.
gingivalis, a purification of the activity from cell extracts was carried out.
4.2.L Methods
4.2.1.1 Determination of the effect of the cysteine proteinase inhibitor TLCK on
NADH oxidase activity.
As the purification of any protein in P. gingivalis might be facilitated if its
cysteine proteinases are inactivated, the effect of the cysteine proteinase inhibitor
TLCK on the activity of NADH oxidase was evaluated to determine if the inhibitor
could be utilised during the purification process.
For this purpose, P. gingivalis V/50 was grown in batch culture in BHI
medium supplemented with 5 mg L-l of haemin and 0.5 g Ll of cysteine. An
overnight-grown culture was used to inoculate a 2OO mL broth that was incubated
overnight, at 3J"C, under normal atmospheric conditions with the flask lid tightly
closed. When an OD566n. was reached, half of the culture was treated with 5 mM
TLCK (final concentration) and half was left untreated. Cell extracts from TLCK-
treated and untreated cultures were prepared in duplicate as described previously in
section 4.11.1. TLCK was also added at 1 mM final concentration to the treated
cultures during washing in buffer and at the beginning, during and end of
sonication. NADH oxidase activity of TlCK-treated and untreated cell extracts
was assayed following the methods of Higuchi (1992), as previously described.
89 4.2.1.2 Growth conditions and harvesting of cells.
To obtain cells for the purification of NADH oxidase, P. gingivalis V/50 was grown under continuous culture conditions in BM medium (Shah et aL.,I916) I supplemented with 5 mg L of haemin, to achieve haemin-excess, and 0.5 g Ll of cysteine. Growth was initiated anaerobically and after 5 generations the incoming
gas mixture was adjusted to produce an oxygenated atmosphere of Nz/COzlOz
(75:5:10). Cells were collected at steady state and harvested by centrifugation
(6000 x g, 4"C for 30 min). Deposited cells were washed twice with 50 mM
potassium phosphate buffer, pH 7.8, containing 0.1 mM EDTA.
4.2.1.3 Preparation of cell extracts.
The collected cells were re-suspended in an aliquot of the same buffer and
lysed by double passage through a French pressure cell at 1200 Kg cm-2. Unbroken
cells were collected after centrifugation (6000 x g, 4"C for 30 min). The cell
extract was treated with deoxyribonuclease I and ribonuclease A to hydrolyse
nucleic acids. An aliquot of extract was stored at -80"C for further analysis. This
sample was designated "S 1".
4.2.1.4 Protein content and enlyvne activity determinations.
Protein content of the cell extracts was determined at each purification step
with a Coomassie Plus Protein Assay Reagent Kit (Pierce, Rockford, IL) using
bovine serum albumin as a standard. NADH oxidase activity was assayed at each
purification step at 25"C following the methods of Higuchi (1992). NADH oxidase
activity was also assayed in I2Vo polyacrylamide gels after native gel
electrophoresis of cells extracts or fractions from different purification steps.
90 Immediately after electrophoresis, gels were incubated for 5 min in 20 mL of rce- cold 0.1 M KPO4 buffer (pH 7.0) and for a further 5 min in the same buffer at room temperature. After incubation, a freshly made NADH solution was mixed into the gel buffer to a final concentration of 50 pg mL-I. After approximately 5 to
15 min, the gel was removed and placed on a UV light box to detect bands of
NADH oxidase activity that appeared as dark bands in contrast to the fluorescent
gel.
4.2. 1 . 5 Ammonium sulphate fractionation.
The cell extract was fractionated by ammonium sulphate precipitation from
20Vo to 75Vo saturation in 5 steps. Initially, ammonium sulphate was added to 2OVo
saturation (0.126 g mLt¡ and the solution was stirred for 60 min at 4'C. The
solution was then centrifuged at 6,000 x g for 30 min and the recovered pellet
dissolved in distilled water. Aliquots of both supernatant and dissolved pellet were
retained for determination of NADH oxidase activity. This procedure was repeated
for 407o, 5OVo, 6OVo and75%o levels of ammonium sulphate saturation.
The dissolved precipitate containing the majority of enzyme activity was
desalted against distilled water using a 30 kda cut-off Centriprep concentrator
(Amicon Inc, Beverly, MA, USA). An aliquot of this sample, designated "S2", was
stored at -80"C for SDS PAGE analysis.
4.2. 1.6 lon Exchange Chromatography (lEC).
An Econo-Pac Q ion exchange cartridge (Bio-Rad), bed volume 5 mL, was
equilibrated with 25 mM Tris, pH 8.0 (Buffer A). The material obtained following
ammonium sulphate precipitation was equilibrated with buffer A using a
9l Centriprep concentrator and 1 mL aliquots of this sample were subsequently applied to the column. Elution of protein from the column, which commenced after the passage of 10 mL of buffer A through the column at a flow rate of 1 mL min-r, was achieved by the application of a linear gradient to a final concentration of 1 M
NaCl (in 25 mM Tris, pH 8.0) at the same flow rate. One mL fractions were then collected and tested for NADH oxidase activity. Fractions exhibiting the highest activity were pooled, desalted and concentrated. An aliquot of this sample, designated "S3", was stored at -80"C for further analysis.
4.2. 1.7 Hydrophobic Interaction Chromatography (HIC).
An Econo-Pac Methyl HIC caftridge, bed volume 5 mL (Bio-Rad) was equilibrated with 1.5 M ammonium sulphate in 0.1 M sodium phosphate buffer, pH
6.8 (buffer A). The active fraction following IEC was equilibrated with buffer A using a Centriprep concentrator and 1 mL aliquots of this sample were subsequently applied to the column. After sample loading, the column was eluted with a linear decreasing gradient of ammonium sulphate, from 1.5 M to 0 M in 0.1
M sodium phosphate buffer, pH 6.8, over 60 min at a flow rate of 1 mL min-I. I mL fractions were collected and tested for NADH oxidase activity. Fractions exhibiting the highest activity were pooled, desalted and concentrated. This final
sample was designated "S4".
4.2.1.8 Sodium dodecyl sulphate polyacrylamide gel electroplnresis (SDS PAGE).
SDS/PAGE was used to monitor purification of NADH oxidase and
estimate the molecular mass of the enzyme. Electrophoresis was conducted in l2Vo
polyacrylamide gels using the standard Tris/glycine buffer system of Laemmli
92 (1970). Gels were run at 125 V until the tracking dye reached the bottom of the gel and were then stained with Coomassie brilliant blue or with the SilverXpressrM Kit
(Novex), according to the manufacturer's instructions.
4.2. 1.9 Molecular mass estimatiott.
To estimate the molecular weight of proteins in SDS gels, a standard curve was constructed with calibration proteins (Molecular weight standards, Invitrogen).
The logarithm of the molecular weight of each standard was plotted against the Rf
value (ratio of migration length/total length of the gel; abscissa). The molecular
weight of the samples was obtained by comparison with the linear calibration
curve constructed.
4.2.1.10 Analysis of sample 54.
Apart from determining the NADH oxidase activity in sample 54
(according to the standard cuvette assay), this sample was analysed for NADH
oxidase activity in the presence of 0.2 mM FAD, NADPH oxidase activity and
production of HzOz during the NADH oxidase assay. Hydrogen peroxide, a
potential product of NADH oxidation, was assayed by a modification of the
method described by Gibson et al. (2000). Briefly, NADH oxidase reactions (total
volume 800 pL) containing 28 nmol NADH, 0.1 M sodium phosphate buffer, pH
J .0, and 93 ttg of sample 54 were allowed to proceed until NADH was completely
oxidized (OD¡¿o > 0.03). Next, 2OO ¡tL of a solution consisting of 2,2'-azino-bis(3-
ethylbenzothiazoline-6-sulphonic acid) at 3 mg ml--l and horseradish peroxidase at
0.2 mg mL-1, in the same buffer, were added to the reaction mixtures. The reaction
was allowed to proceed for 20 min at room temperature and then the 4566n. wâs
93 measured. Samples were compared to a standard curve generated by known concentrations of HzOz. Data presented represent the mean of samples assayed in triplicate in two independent experiments. As a control, a known concentration of
H2O2 was added to the reaction mixture after completion of the NADH oxidase reaction and before adding the 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) and the horseradish peroxidase.
4.2.1.11 ldentffication of the protein with NADH oxidase activity in sample 54.
As the SDS polyacrylamide gel electrophoresis showed that the final fraction (S4) contained more than one band, then, in order to identify which of them was responsible for the NADH oxidase activity, sample 54 was run in a native gel and assayed for NADH oxidase. Analysis of this gel under UV light showed two bands presenting activity. The R¡ of these bands was measured and the gel was then blotted onto a nitrocellulose membrane. The membrane was stained with Coommasie blue and the bands with NADH oxidase activity, identified according to their R¡ value, were excised from the membrane. The protein was eluted from the membrane pieces by placement overnight in a 50Vo acetonitrile solution. Elution was verified by checking that there was no Coomassie blue dye present in the membrane, while the supernatants were blue in colour. The supernatants, containing the eluted protein from each gel piece, were then concentrated in a Jouan RC10.10 rotary evaporator, redissolved and denatured in
SDS sample buffer, and re-run in an SDS gel together with sample 54. The bands containing the NADH oxidase activity were then compared to the bands in the 54 fraction, in order to identify, in this fraction, which bands were the putative NADH oxidases.
94 4.2.1.12 Mass spectrometry (MS) identification of purified protein.
MS identification of the purified protein was performed in Dr. Gary
Corthals' Mass Spectrometry and Protein Analysis Laboratory at the Garvan
Institute of Medical Research (Sydney, NSV/). Proteins were identified using a tandem mass spectrometer (MS/MS) coupled to an in-line microcapillary liquid chromatography system (pLC). Briefly, the method consisted of the extraction of the gel-separated proteins from the gel matrix by digestion with trypsin. The peptides were then extracted and concentlated by pLC. Each eluting peptide was fragmented by the mass spectrometer in a data-dependent manner. The amino-acid sequence of the different peptides was derived from the fragmentation pattern and the resulting sequence correlated with the P. gingivalis genome database
(http://www.tigr.org) translated into protein, to identify the ORF corresponding to the protein from which each peptide was generated.
4.2.2 Results and discussion
4.2.2.1 Determination of the effect of the cysteine proteinase inhibitor TLCK on
NADH oxidase activity.
NADH oxidase activity in cell extracts was reduced by more than half when TLCK was added during the cell extract preparation. Although this represented an obstacle - in that the absence of TLCK could lead to proteolysis of the enzyme while its presence leads to its inactivation - it was decided to purify the enzyme without adding TLCK, to avoid any interference with the catalytic properties.
95 4.2.2.2 NADH oxidase purirtcafion
Ammonium sulphate fractionation of sample 51 (extract of broken cells) resulted in the majority of the activity concentrated in the 6OVo ammonium sulphate precipitate. However, activity was also found in other precipitates as shown in Table 4.3.
Table 4.3. NADH oxidase activity after ammonium sulphate
fractionation of sample 51
Ammonium sulphate NADH oxidase activity (units
concentration (dv) mg protein-l)u
20% 0
40Vo 2.8
50Vo 5.4
60Vo 18.1
15Vo 5.6
o One unit of enzyme activity was defined as the amount of extract that
catalysed the oxidation of I nmol of NADH min-r.
ln order to verify that the same protein was responsible for the activity in all the precipitates, a native polyacrylamide gel was run with 100 ug of each sample (4OVo, 5OVo, 60Vo and 75Vo precipitates) and the activity of NADH oxidase was assayed in the gels after electrophoresis. That this native gel showed one single band per sample, running at a similar distance and containing NADH
96 oxidase activity implied that the same protein was possibly responsible for the
activity in all fractions.
Thus, the 6OVo ammonium sulphate precipitate, containing the highest
activity, was designated sample 52 and used for subsequent IEC separation.
Figure 4.4 shows the results of a typical separation of sample 52 in the IEC
system. After several separations, samples containing NADH oxidase activity were
desalted and concentrated and this sample, designated "S3", was further separated by HIC.
Figure 4.4. Purification of NADH oxidase by ion exchange
chromatography
16 14 12 10 NADH I oxidase -MDt-loxidase actlvlty 6 ac-tivity 4 (Lhits/nl) 2 0 1 3 5 7 I 11 1315171921æ25
Fraction number
97 Figure 4.5 shows the results of a typical separation of sample 53 in the HIC
system. After two separations, samples containing NADH oxidase activity were
desalted and concentrated and this sample was designated 54.
Figure 4.5 Purification of NADH oxidase by hydrophobic interaction
chromatography
35
30
25
20
15
10
5
0 0 2 4 6 8 1012141618202224262830
Fraction number
Figure 4.6 shows a SDS polyacrylamide gel of samples from each purification step. 100 pg of sample 51, 52 and 53 and2O pg of sample 54 were loaded in the gel. This gel was sliced in two after electrophoresis and half was
stained with Coomassie Blue (for samples Sl, 52 and S3), while the other half was
silver stained (for sample S4). This procedure was adopted in order to minimise the
amount of sample 54 to be utilised, as silver staining is 100 times more sensitive than Coomassie blue staining.
98 Table 4.4 shows the NADH oxidase activity in each of the purification fractions
Figure 4.6. SDS polyacrylamyde gel of fractions from each purification step.
Protein Protein kDa standards 51 52 53 S4 standards
200 200 ffi 116 r16
91 9',7
66 66
55 ,#,æ"'" 55
37 37
3t 3l
21 21
Table 4.4. NADH oxidase activity in purification fractions.
Fraction NADH oxidase activity (Units
mg protein-l)
S 1 (crude cell lysate) 15.8
32 (6OVo (NH¿)zSO¿ precipitate) 18.7
53 (IEC fraction) 87.2
54 (HIC fraction) r49.2
99 4.2.2.3 ldentification of the protein with NADH oxidase activity in sample 54.
As shown in Figure 4.6, the SDS PAGE analysis of the purified fraction
(S4) showed two bands of 54.19 kDa and 45.28 kDa and although they constituted most of the protein in the sample, other bands in lower concentrations were also present. Therefore, to identify the band in the SDS polyacrylamide gel responsible for the NADH oxidase activity, sample 54 was run in a native gel and assayed for
NADH oxidase. Analysis of this gel under UV light showed two bands displaying activity. Figure 4.7 shows the SDS gel where the two bands were re-run after being excised from the native gel and compared to the 54 fraction.
Figure 4.7. SDS silver stained gel of sample 54 and eluted bands
presenting NADH oxidase activity.
Sample 54 Track lo 'lrack2 protein standards kDa
66
55 Band B¡
Band C1
JI
31
21 o Tracks I and 2 correspond to the bands containing NADH oxidase
activity, excised from the native gel, and re-run in this SDS gel.
100 As can be seen from the gel in Figure 4.J, track 1 contained two bands corresponding to the major bands in sample 54; that is, bands B and C. In contrast,
frack2 showed two bands, one of them corresponding to band B and another band, designated band A, that is not represented in sample 54. There are two possibilities for the appearance of this band in this gel. The first is that the protein in the re-run
sample was incompletely denatured and migrated at a different molecular weight, although band A contains, in fact, the same protein as band B. The second option is that band A contains a different protein to that in band B, not present in sample 54 in sufficient quantity to be visualised by the staining procedures (as this band comes from a native gel, where greater amounts of sample were loaded than in the
SDS silver-stained gel). Thus, to clarify these results, all bands (4, B and C) from
sample 54 and both tracks were included in the mass spectrometry analysis.
4.2.2.4 Mas s spectrometry identffication
The identification data for bands B1 and C1 (from sample S4), Bz and Cz
(track 1) and A and B3 (track 2), shown in Figure 4.7, are presented in Table 4.5.
101 Table 4.5. Mass spectrometry identification data.
Sample M\ry MW Peptides Sequenced Protein ID" observedu expectedb
Band A 64.5 61.5 (R)AEQQKKEIIGK PG 1140
Band Br 54.2 54.0 (R)HPETGPTVK PG 0625 (R)IENPVDHPMIR (R)LAEMDEYK
Band Bz 54.2 54.0 (R)IENPVDHPMIR PG 0625
Band Be 54.2 54.0 (R)LAEMDEYK PG 0625
Band Cr 45.3 54.0 (R)YLAGATGK PG 0625 (R)HPETGPTVK
Band Cz 45.3 54.0 (R)YLAGATGK PG 0625 (R)HPEIGPIVK
u Observed molecular weight calculated according to protein standards.
b Expected molecular weight as calculated from the P. gingivalis'W83 nucleotide
sequence (http ://www.oralgen.lanl. gov).
" Accession numbers correspond to the numbers from the Oral Pathogen Sequence
Database (http ://www.oralgen.lanl.gov).
Band A was identified as PG 1140 in the P. gingivalis nucleotide database; an ORF that has been annotated as a putative ribosomal protein. Thus, it was confirmed that this protein (Band A) migrated together in the native gel with the protein in Band B and the two proteins were only separated in the SDS gel. This is not an unexpected result as the resolving power of a native polyacrylamide gel is not very high, the proteins not being denatured. This band, however, being a ribosomal protein, is not likely to posses any NADH oxidase activity.
Band B was identified as PG 0625; an ORF annotated in the database as 4- hydroxyphenylacetate-3-hydroxylase (Hpa). However the annotation is confusing
102 as the gene name given to this protein is abfD, which is the gene symbol for a different protein, the enzyme 4-hydroxybutyryl-CoA dehydratase (AbfD). The annotated database reports that this ORF presents l5%o similarity to a 4- hydroxybutyryl-CoA dehydratase in Clostridium aminobutyricum (CA860035)
and 58Vo, 38Vo and 34Vo simllarity to three 4-hydroxyphenylacetate-3-hydroxylase proteins in Archaeglobus fulgidøs (NP_069169.1; NP_069860 and NP_069718).
Band C was identified also as PG 0625. Therefore, band C is a fragment of the same protein as in band B, probably degraded during the purification procedures. Fragmentation most probably occurred during the HIC step as the HIC purification diagram (Fig. a.5) shows two peaks that indicate the presence of two proteins with NADH oxidase activity; although they were eluted from the column at slightly different times, they were pooled together in the same sample (S4).
Therefole, it appears that the ORF designated PG0625 coresponds to the protein containing NADH oxidase activity, purified in these experiments.
These results, including the ORF identity and its NADH oxidase activity, will be further analysed and discussed in section 4.3 of the present chapter.
4.2.2.5 Analysis of NADH oxidase activity in sample 54
As shown in Fig. 4.8, the presence of 0.2 mM FAD increased the NADH oxidase activity in sample 54 by I3Vo, and would, therefore, appear to act as a co- factor of the protein responsible for the NADH oxidase activity in the purified sample. It was also observed that, in the presence of NADPH, the purified protein had no oxidase activity (Fig. a.9). These results agree with previous observations with the cell extracts.
103 Table 4.6 shows that HzOz was produced rather than consumed during the
NADH oxidase assay. These assays indicated that the purified protein produces
H2O2 during the NADH oxidation and does not have any peroxidase activity. On average, 1 mol of HzOz was detected per 2.8 mols of NADH added to the reaction.
The detection of H2O2 as a product of the NADH oxidase activity present in this protein does not agree, however, with some of the results presented previously in
Section 4.2.2.3. These results suggested that, as catalase did not have any effect on the NADH oxidase activity present in cell extracts, the enzyme did not produce any significant amounts of HzOz.An explanation for these contradictory results will be presented at the end of this chapter (Section 4.6).
104 Figure 4.8. NADH oxidase activity in sample S4u in the presence and
absence of FAD (0.2 mM).
0.48 0.46 0.44 o.42 A¡+o 0.4 o NADH oxidase 0.38 NADH oxidase 0.36 n + FAD 0.34 0.32 0.3 0 2 4 6
Time (min)
u 50 ug of protein were added to the reaction mixtures
Figure 4.9. Anatysis of NADPH oxidase activity in sample S4u.
0.48 0.46 0.44 e NADH oxidase 0.42
A¡¿o 0.4 l NADPH oxidase 0.38 0.36 0.34 0.32 0.3 0 2 4 6 8 Time (min)
u 50 ug of protein were added to the reaction mixtures.
105 Table 4.6.HzOz produced by the NADH oxidase reaction.
Reaction Mixture ( lml-)" H2O2 detected (nmol)
NADH oxidase assay 11 I 1 + 1 .2
NADH oxidase assay + 5 nmol HzOz 14.01+0.8
NADH oxidase assay + 20 nmol HzOz 28.20+t.6
o NADH oxidase assay contained 93 ug of protein from sample 54 and
28 nmol NADH. H2O2 was detected after NADH consumption was
completed. 5 nmol and 20 nmol of H2O2 were added to completed
NADH oxidase reaction mixtures as a control.
4.3 Analysis of PG 0625 sequence
4.3.1 Gene ID
PG 0625 was the sequence identified in P. gingivalis genome corresponding to the purified NADH oxidase. According to the annotated database's (http://www.oralgen.lanl.gov) sequence similarity data, this ORF could encode for a 4-hydroxybutyryl-CoA dehydratase (AbfD) (most likely match) or a
4-hydroxyphenylacetate-3-hydroxylase (HpaA). To understand the mechanism by which this protein shows NADH oxidase activity, the identity of the enzyme must be first established. Unfortunately, the poor yield obtained during the purification procedures did not provide enough enzyme preparation to directly assay its activity. However, the sequence of the enzyme was analysed to establish the reason
106 for the ambiguous annotation in the database and to assign the ORF its most likely identity.
4.3.1.1 Merhods
All alignments were performed using the program T-coffee
(http ://www. ch. embnet. orf/software/tcoffee. html). PG 0625 sequence was obtained from the P. gingivalis genome database (http://www.oralgen.lanl.gov), Clo stridium dfficile putative AbfD sequence was obtained from the unfinished genome database (http://www.sanger.ac.uk/projects/c_difficile/). The rest of the sequences were obtained from the NCBI Entrez database (http://www.ncbi.nlm.nih.gov).
PG 0625 sequence was aligned with two 4-hydroxybutyryl-CoA dehydratase (AbfDs) sequences from Clostridium aminobutyricum (CA860035.1) and Clostridium dfficile. The AbfD from Clostridium aminobutyricunt has been isolated and its properties characterised (Scherf and Buckel, 1993), while the
Clostridium dfficile sequence is a putative AbfD identified by sequence similarity
(Gerhardt et al., 2000).
PG 0625 was aligned also with two 4-hydroxyphenylacetate 3- hydroxylases, large component (HpaB), belonging to Escherichia coli (Q57160)
(Prieto and Garcia, 1994) and Klebsiella pneumoniae (AAC37I20) (Gibello et al.,
1997) which properties have been estabiished after purification.
The third alignment presented shows PG 0625 and A. fulgidus putative Hpa
(NP_069169.1) that was identified as such by sequence similarity (Klenk et al.,
1991). The fourth alignment is for A. fulgidus putative Hpa and C. aminobutyricum
AbfD.
101 4.3.1.2 Results
Figure 4.10 shows the alignment of PG 0625 and the two AbfD from
Clostridium atninobutyricum and Clostridium dfficile. It can be seen from the alignment that the similarity among the three sequences is remarkably high, while
PG 0625 does not exhibit the same level of similarity when compared to
Escherichia coli and Klebsiella pneumoniae HpaB (Figure 4.II).
However, PG0625 identity was annotated as a possible HpaB based on a putative Hpa from A. fulgidus and, indeed, the sequences show a high level of homology (Figure 4.12). Therefore, it is likely that this ORF in A. fulgidus, wrongly annotated as an Hpa, could be an AbfD. This assumption was verified in the alignment shown in Figure 4.13 betweenA. fulgidus putative Hpa and the C. aminob uty r ic um well characterised AbfD.
It is worth noting that, althoughA. fulgidzs sequence name is HpaA-l, the component of the Hpa enzyme to which it has more similarity is component B, the large oxygenase component (HpaB). Component A is a possible regulatory protein of the enzyme complex that is composed of only 268 residues; as opposed to the
=500 residues of component B (Prieto and Garcia, 1994). No ORF encoding for component A was found in P. gingivalis genome.
108 Figure 4.10. Sequence alignment of PG 0625, 4-hydroxybutyryl-CoA
dehydratase (AbfD) fromClostridium øminobutyricum and a putative AbfD
fr om C I o s tridium dffi c il e"
C. ambut MLMTAEQY I E S LRKLNTRVYMFGEKI ENWVDHPM] RPS INCVRMTYELAQDPQYA. C . diffic - LMTGAQYT E S LRKLNTKVYMFGEEVKN!ÙVDHPMT RPS INCVAATYDLAHDPEYA. PG1625 -MMT SEQYVE S LRKLNLKVYFMGERIENPVDHPMI RP SMNSVA}4TYKLAEMDEYK :**, **.******* ,**..**...* *********.*.* **.**. :* YY C. ambut KSNLTGKTINRFANLHQST D DLRKKVKMQRL LGQKTAS C FQRCVGMDAFNA VFST' c . diffic TSNI TGEKINRFGHLHQSVDDL T KKVKMQRLCGQKTASCFQRCVGMDAFNAVYS T' PG0625 TSNLTGKQVNRFCHLHQS**. *. .*** .**** TEDLKDKVKMQRLMGQKTASCFQRCVGMDAFNAIYS .** ******* *******************'.** T'
C. ambut QKYGTNYHKNFTEYLKY I QENDL T VDGAMT DPKGDRGLAPSAQKDPGLFLRIVEK: C.diffic KAHGTNYHDNFVKYLTY T QENDLVVDGAMT DPKGDRS L S PSAQPDPDMFLH IVER. PG1625 EYMKYVQDNDLVVDGAMT DPKGDRGL S PSEQADPDLYLH IVEV: QALGTTYHKRFT: **.**..* :*:.*.*,***'************.*.** * **,::*.***
C. ambut VVRGAKAHQTGS INSHEH I I MPT TAMTEADKDYAVS FACPS DADGLFMI YGROSY C.diffic IVRGAKAHQTGS TNSHEHLTMPT I SMTEADKDYAVS FAVPS DAEGVFMT YGRQSC. PG]625 VVSGAKAHQTGAVNSHEHLIMPTIAMREADADYAVSFAVPS.* ********..*****.*****.* *** ******* ****'*..******** DAEGVIMTYGRQSC. Y C. ambut EE GAD 1 DIGNKQFGGQEALVVFDNVFT PNDRI FLCQEYDFAGMMVERFAGYHROS C.diffi-c EE GADVDLGNKE FGGQEALVVFDNVFVPNDRI FLCGEI/ÙDFS GMLVERFAGYHRQS PG0 62 5 EE GADI DLGNSE FGGHEALVVFDRVFVPNERVFMCKEYQFAGMMVERFAGYHRQS *****. ****. . ***. *******. **. **. *. *. * *: : *: **. *********** YY C. ambut VGVGDVVI GAAALAADYNGAQKASHVKDKL I EMTHLNETLYCCG]ACSAEGYPTA" C. dÍffic VGVGDV I I GAAÄLAADYNGANKASH I KDKL I EMTHLNE S LYCCG IACS SEGHKTE, PG0 62 5 VGVG******. DVL T************. GA.AALAADYNGVPKAS ****. H I KDKL******* T EMI HLNETLYACGIACS****. **. ******. SEGTQMK, ** Y C. ambut I DLLLANVOKON T TRFPYE IVRLAE DTAGGLMVTMPSEADFKSEÎWGRDGET I G. C . diffic T DLLLANVCKQNVTRFPYE IVRLAEDIAGGLMVTMPSEKDFKS DLKVGTSGMT ] G: PG0 62 5 I DLLLANVCKQN I TRLPYE ]ARLAEDIAGGLMVTMPSQQDFRHP- - _ - T G. ************. **. ****. ****************. **. - - - -E **
C. ambut FFAÄAPTCTTEERMRVLRFLENI CLGASAVGYRTESMHGAGS PQAQRIMIARQGN C. diffic YFKAS SVAS TEERMRI LRFLENI CLGS SAVGYRTE SMHGAG S PQAQRIMI SRQGN PG0625 YLAGATGKS TENRMRVLRL TEN I TLGTAAVGYRTESMHGAGS PQAQR]MI ARQGD: **. ***. **. . *** ** . . ********************** . )k**
C, ambut ELAKAIAGIK------C . diffic ELAKKIAGIKKEEALN PG]625 KLARATAHIDESLDK- :**: ** *.
u Arrows indicate conserved cysteine residues in the three sequences
109 Figure 4.11. Sequence alignment of PG 0625 and 4-
hydroxyphenylacetate 3-hydroxylase (oxygenase sub-unit) (HpaB) from
Escherichía coli and. Klebsiella pneumoniae.
E. coli MKPE DFRASTORPFTGEEYLKSLQD-GRE I YI YGERVKDVTTHPAFRNA.AASVAQ. K. pne MKPENFRADTKRPLTGEEYLKSLQD-GRE T YI YGERVKDVTTHPAFRNAAASVAQ, PG0 62 5 - -MMTSEQYVE.* *.*..**. SLRKLNLKVYFMGERIENPVDHPMTRPSMNSVAM' ..*' ***... ** .* . ***
E . coli HKPEMQDSLCWNTDTGSGGYTHKFFRVAKSADDLRHERDAIAEWSR],SYGI'ùMGRT K. pne HNPELQNTLCWGTDTGSGGYTHKFERVAKSADD],RQQRDAIAEWSRLSYGWMGRT PGO 62 5 EMDEYKH-LMTATSNLTGKQVNRFCHI,HQS TEDLKDKVKMQRLMGQKTASCFQRC' * . * * .* ..* .*..**. ' . *
E. coli YKAAFGCALGGTPGFYGQFEQNARNWYTRI QETGLYFNHAIVNPP T DRHLPTDKV: K. pne YKAAFGGGLGANPGEYGQFEQNARDWYTRI QETGLYFNHAIVNPP I DRHKPADEV: Pc0625 FNAI..* YSTTYEMDQALGTTYHKRF IEYMKYVQDNDLVVDGAMTDPKGDRGLS.*. * . *. .* ** PSEQ,
E . cofi Y I KLEKET DAGI IVSGAKVVATNSALTHYNMI GFGSAQVMGENPDFALMFVAPMD, K. pne Y I KLEKET DAG I IVSGAKVVATNSALTHYNMI GFGSAQVMGENPDFALMFVAPMD, PG1625 YLH IVEVRE DG IVVSGAKAHQTGAVNS HEHI' I -MPT ]AMREADADYAVS FAVPSD, *::: : : **.*****. *.:. :* ::* : : : :.*.*. *..* *
E. coli L I - SRASYEI4VAGATGSPYDYPLSSRFDENDAI LVMDNVLI PWENVLLYRDFDRC. K. pne L ] - SRASYELVAGATGS PYDYPLSSRFDENDAI LVMDNVL I PWENVLlYRDFDRC. PGo 62 5 M1.* YGRQSCDTRKMEEGADI * * . *. DLG-NSE* * FGGHEALVVFDRVFVPNERVFMCKEYQFAI* ..*..*.* *..* * *..
E . cofi E GGFARMY PLQACVRLAVKLDF I TALLKKS LE CTGTLE FRGVQADLGEVVAWRNT. K. pne EGGFARMYPLQACVRLAVKLDFI TALLKRSLECTGTLE FRGVQAELGEWAWRNM. PG]625 RFAGYHRQSYGGC -KVGVG_ DVLI GAAALAADYNGVPKASHIKDKL IEMI HLNET. : . ,* ::.* *.: , : : .*. : :: .* *:: .:
E. cofi DSMC SEATPWVNGAYLPDHAALQTYRVLAPMAYAKIKN T TERNVT SGL I YLPS SA: K. pne DSMCAEATPWVNGAYL PDHAALQTYRVMAPMPYAKTKN I IERSVTSGLI YLPS SA. PG0625 ]ACS SEGTQMKAGNYMI DLLLANVCKQNI TRLPYE IARLAE - DIAGGLMVTMPSQI . ,* * * *. * .* . * **. *
E. co 1i PQI DQYLAKYVRGSNGMDHVQRIKI LKLMWDAI GSE FGGRHELYE TNYSGSQDE I: K.pn e PQINDTLAKYVRGSNGMDHVERI KI LKLMWDAI GSE FGGCHELYE INYSGSQDE I. I EN I TLGTAAVGYRTESMHGAGS PQAQ. PGO 6 25 PE*.* I GP IVKKYLAGATGKSTENRMRVLRL' **. *. * .*...*.*. .**
E . coli RQAQS S GNMDKMMAMVDRCL SE YDQN GI/ÙTVPH LHNN DD I NMLDKLLK K. pne ROAQ S S GNM DKMMAMV DRC L S E Y DQN GWTV P H L HNN T D I NML DKL LK PG1625 RQGDLEGK-----KKLARATAHI DES - --- -LDK- - - **.: .*: I *.::. *1. ***
110 Figure 4.12. Sequence alignment of PG 0625 and a putative 4-
hydroxyphenylacetate 3-hydroxylase (oxygenase sub-unit - HpaA-l) from
A. fulgidus.
A. fulg MEGMMNGKEY IE SLRKÍ,KPKVYFMGRKT DSVADDPI lAPHVNAAAMTYELANDPR PG1625 - - -MMTSEQYVE SLRKf,NLKVYFMGERT ENPVDHPMIRPSMNSVAMTYKLAEMDE ** ..*.******. ****** .*. * *' * '*' ****!**'
A. fulq TAT S HLTGEKINRFTH I HQSTE DLVKKVRMMRLLGRKTGSCFQRCVGLDALCALY Y PG0 62 5 TAT****.***...*** SNLTGKQVNRFCHLHQS *.******* TE DLKDKVKMQRLMGQKTASCFQRCVGMDAFNAI .**'* **.*.**.********.**. *:*
A. fulg VDRKYGT DYHERFKDYL I HVQKNDLMAAGAMTDVKGDRSLRPHQQADPDVYVRVV. PGO 62 5 MDQALGTTYHKRFI.*. ** **.** EYMKYVQDNDLWDGAMTDPKGDRGLS.*. .** ***. ***** **** * PSEQADPDIYLH* .*****.*...* IV-
A. fulg G IVVRGAKAH I TGAVN SHE I IALPTRAMGEE DKDYAVAFAVPVDAEGVVMI FGRQ' PG0625 G**** TVVSGAKAHQTGAVN ***** ******** SHEHL . IMPT.** IA¡4READADYAVS ** * * ****.**** FAVP S DAEGVTMT*****.**.*** YGRQ
A. fulg RLDGDI DCGNAKYATVGGEALI I FNDVFVPt/ÙERVFMCGEYDFASELVETFATYHRI KMEEGAD I DLGNSE FGGHEALVVFDRVFVPNERVFMCKEYQFAGMMVERFAGYHRI PG0 62 5 ::: . .: * ***..*. **** ****** **.**. .** ** ***
A. fulg CKVGVADVL I GASAAIAEYNGVERASH IVDKLTEMVHLAETCWCC SLACSYEGFK' IEMI HLNETLYACG IAC S SEGTQ] PG0625 CKVGVGDVL***** ******.* I GA.AALAADYNGVPKASH *.**** .**** IKDKL *** **.** ** . * .*** **
A. fulg YMPNLLLANVTKLN 1 TRFPYEWSRLAQDIVGGLVATT PSELDYKNPETRDLIEKY I GP TVKKY PG0 62 5 YMI** DLLLANVCKQN.****** * ****.*** I TRLPYE TARLAE .***.**.***..* D]AGGLMVTMPSQQDERHPE **. *...** :::**
A. fulg DVPTEHRVRMTRLIENLS I GAEL- - -- PE SMHGAGS PAAQKIMIARRAN I DVKKE. Pc0625 GKSTENRMRVLRL I EN I TLGTAAVGYRTE SMHGAGS PQAQRIMIARQGDLEGKKK. . .**:*.*..*****...*. ********* **'*****', " ' **:
A. fulg AG I LPDE DYLRVRGFKREGDEGRS PG0 62 5 AHI*** DESLDK-
1lr Figure 4.13. Sequence alignment of a putative 4-hydroxyphenylacetate
3-hydroxylase (oxygenase sub-unit - HpaA-l) fromÁ. fulgídus and 4-
hydroxybutyryl-CoA dehydratase (AbfD) from Clostrídium
aminobutyricum
hpaA-1,/4. fulg MEGMMNGKEYIESLRKLKPKVYFMGRKI DSVADDP I TAPHVNAAAMTYELANDPR AbfDC. ambut - -MLMTAEQYIESLRKLNTRVYMFGEKIENWVDHPMIRPS,* ..********, ,**..* **. * *. * TNCVRMTYELAODPQ.* ******.**. hpaA-1/4. fulg TATSHLTGEKINRFTH ] HQS TEDLVKKVRMMRLLGRKTG SCFQRCVGLDALCALY AbfDC. ambut TTKSNL*..*.* I GKT*..****...****,** TNRFANLHQS TDDLRKKVKMQRLLGQKTASCFQRCVGMDAFNAVF ***,* ****.**.********.**. *.. hpaA-1,/,A. fulg VDRKYGTDYHERFKDYLI HVQKNDLMAAGAMTDVKGDRS LRPHQQADPDVYVRVV. AbfDC. DQKYGTNYHKNFTEYLKY IQENDL IVDGAMTDPKGDRGLAPSAQKDPGLFLRÏV. ambut I.*.****,**, * .** '.*.***. ***** **** * * * ** ...*.* hpaA-1,/4. fulg GIVVRGAKAHITGAVNSHE I IALPTRAMGEEDKDYAVAFAVPVDAEGVVMI FGRQ' AbfDC. ambut G********** IWRGAKAHQTGS **..**** TN SHEH *I IMPT .** TAI"ITEADKDYAVS ** * ******.** FACPS * DADGLFMI**.*. **.*** YGRQ, hpaA-1 /.A'. fulg RLDGDI DCGNAKYATVGGEAL I T FNDVFVPI/{ERVFMCGEYDFASELVETFATYHRi AbfDC. ambut KMEEGADI DLGNKQFGGQEALVVFDNVFT PNDRI FLCOEYDFAGMMVERFAGYHRI ::: . .: * ***..*..**.* .*.*.* *****. .** ** *** hpaA-1,/4. fulg CKVGVADVL I GASAAIAEYNGVERASH IVDKLTEMVHLAETCWCCSLAC SYEGFK' AbfDC. ambut CKVGVGDVVI***** **.***.* GAAALAADYNGAQKASHVKDK], *.*** ,.***. *** lEMTHLNETLYCCG ** ** ** .** TAC.*** SAEGYP' **. hpaA- 1,/4. f ulg YMPNLLLANVTKLNI TRFPYEWSRLAQDIVGGLVATTPSELDYKNPET- - AbfDC. ambut YQT* DLLLANVCKQNT.****** * ******** TRFPYE IVRLAEDIAGGLMVTMPSEADFKSETWGRDGET ***.**.***:.* *** *.*. hpaA-1,/4. fulg EKYLKGKADVPTEHRVRMI RL IENLS I GAEL- - - - PE SMHGAG S PAAQKTMIARR, AbfDC. ambut NKFFAÄAPTCTTEERMRVLRFLEN I CLGASAVGYRTE SMHGAG S POAQRIMIARQI :*:: .**.*:*::*::**...** ********* **.*****. hpaA-1/4. fulg KKEEAKRVAG ILPDE DYLRVRGFKREGDEGRS AbfDC. ambut KKELAKAIAGIK----.---*** ** .***
112 Based on the sequence alignments it is, therefore, likely that PG 0625 has identity with an AbfD. Further evidence to support this will be presented and discussed below.
4-hydroxybutyryl-CoA dehydratase (AbfD) is an enzyme that has a possible role in the glutamate fermentation pathway of P. gingivalis.Takahashi ¿r aL (2000) showed that P. gingivalis cell extracts possessed most of the activities involved in this pathway (see Figure 4.I4).
AbfD, in particular, is involved in the dehydration of hydroxybutyryl-CoA to crotonyl-CoA, and although this step does not appear in Takahashi's proposed pathway, other studies (Scherf and Buckel, 1993; Gerhardt et a1.,2000; Buckel,
2001) in Clostridium aminobutyricum, have shown that it is an essential step in the conversion of 4-hydroxybutyrate to crotonyl-CoA. This additional step is also included in Figure 4.14.
113 Figure 4.l4.Metabolic pathway for glutamate as proposed by
Takahashi et al. (20O0), with a modification of the 4-hydroxybutyrate- crotonyl-CoA step as proposed in the present study. The relevant enzymes are indicated in the grey boxes.
Glutamate NAD NH¡ -+< NADH 2-oxoglutarate
CoA Coz 2H
Succinyl-CoA
NAD CoA NADH
Succinate-semialdehyde
NAD(P)H I NAD(P) 4-hydroxybutyrate
/ Hzo Acetyl-CoA 4-hydroxybutyryl-CoA
Acetate
\ 7 Crotonyl-CoA 2H
Butyryl-CoA
Acetate I Acetyl-CoA Butyrate
tt4 Recently, Gerhardt et al. (2000) found that the AbfD gene of C. aminobutyricum forms part of a genetic region containing other genes involved in the same fermentation pathway. The affangement of these genes is surprisingly similar in P. gingivalis (see Figure 4.15) and provides further evidence to support the idea that PG 0625 is indeed an AbfD.
Figure 4.15. Genetic arrangement of the region containing PG 0625
(AbfD) and other genes involved in the glutamate fermentation pathway of P.
gingivalis compared to the same region in C. aminobutyricum (for
abbreviation of enzymes and genes see Fig. 4.14).
C. aminobutyricum
AbfT 4TIbD
P. gingivalís
AbfT 4HbD
115 On the other hand, the possibility that PG 0625 corresponds to a 4- hydroxyphenylacetate-3-hydroxylase is less likely. No metabolic pathway involving this enzyme has been suggested for P. gingivalis. Moteovet, although 4- hydroxyphenylacetate-3-hydroxylase is a two component flavin monooxygenase which catalyses the hydroxylation of p-hydroxyphenylacetate to 3,4- dihydroxyphenylacetate, using NADH and oxygen in the reaction, the component capable of the NADH oxidase activity is the HpaC component, and not the oxygenase component (HpaB), which is the one similar to PG 0625. No other ORF possibly corresponding to the coupling protein (HpaC), was found in P. gingivalis genome sequence.
Furthermore, the AbfD of C. aminobutyricum exists as a homotetramer containing one FAD and one [4Fe-4S] per subunit. The alignment in Fig.4.10 shows 7 conserved cysteine residues, indicated by the arrows, in the AbfD sequences. It has been demonstrated that some of these cysteines are involved in the formation of the l4Fe-4Sl cluster that, although not contributing to catalytic activity, seems to play an important structural function (Gibello et al., 1997; Müh et al., 1997). On the other hand, the alignment in Fig. 4.11 shows that the mono- oxygenases possess none of these conserved cysteine residues, while HpaA-1 from
A. fulgidus does contain the cysteines, reinforcing its identity as an AbfD.
4.3.2 Further evidence of NADH oxidase activity in an AbfD protein
As no repofi in the literature of an AbfD containing NADH oxidase activity was found, Dr. Wolfgang Buckel (Laboratorium fuer Mikrobiologie, Philipps-
Universitaet, Marburg, Germany), who has studied extensively the AbfD mechanism, was contacted. He carried out assays for NADH oxidase activity in C.
116 aminobutyricum purified AbfD preparations according to the methods of Higuchi
(1992). In a personal communication, Dr. Buckel informed me that he found
NADH oxidase activity in two different preparations of the enzyme that had been kept frozen for about two years. One preparation contained about 10 units mg protein-land the other contained2 units mg protein-I. However, he reported that with NADPH the activities were half those for NADH, contrasting with the results from P. gingivalis AbfD which did not exhibit any activity with NADPH. The mechanism by which AbfD could act as a NADH oxidase and the biological significance, if any, of this reaction will be discussed in Section 4.6.
4.4ldentifÏcation of other candidate genes responsible for NADH oxidase activity in the P. gingivalis genome sequence
The results presented previously indicated that the main NADH oxidase activity present in P gingivalis cell extracts might be the result of the NADH oxidase activity displayed by the 4-hydroxybutyryl-CoA dehydratase (AbfD).
However, as the purification of the enzyme was carried out under conditions where P. gingivalls cysteine proteinases had not been inactivated, the possibility that other proteins contained NADH oxidase activity, but were missed during the purification, was considered. Therefote, the P. gingivalis genome database was searched for other proteins likely to possess NADH oxidase activity.
117 4.4.1Methods
Three sequences were obtained, from different microorganisms, for each of the two well-characterised NADH oxidase systems; the one component H2O- producing NADH oxidase (Nox) and the two-component NADH oxidase/alkyl hydroperoxide reductase system (AhpF-C). The sequences were obtained from the
Swiss-Prot and TTEMBL database accessed through Expasy molecular biology server (http://expasy.ch) (Table 4.7). These sequences were searched, one by one, against the P. gingivalis nucleotide database translated into protein, using Blastp
(Gish, 1996-2000) at The Institute for Genomic Research Comprehensive
Microbial Resource (tigr cmr) (http:lltigrblast.tigr.org). The sequences were then aligned with the identified closest matches in P. gingiualis, using the program T- coffee (http ://www.ch.embnet.orf/software/tcoffee.html).
118 Table 4.7. Sequences obtained from the Swiss-Prot and TTEMBL databases used to identify NADH oxidase-like proteins in P. gingivalis.
Protein name Microorganism Accession number Reference
HzO-forming NADH oxidase Streptococcus mutans Q544s3 (Matsumoto et al., 1996)
NADH oxidase Streptococc us .faecalis P3706r (Ross and Claiborne, 1992)
NADH oxidase Lactococcus lactis Q9CDTS (Bolotin et aL.,2001)
NADH oxidase/alkyl hydroperoxide reductase Streptococcus mutans o66266 (Higuchi et aL.,1994)
NADH oxidase/alkyl hydroperoxide reductase Amphibacillus xylanus Q06369 (Niimura et aL.,1993)
Alkyl hydroperoxide reductase subunit F Escherichia coli P35340 (Blattner et aL.,1997)
I l9 4.4.2 Results and discussion
The Blast search indicated that the most likely ORF matching a Nox in P. gingivalis was PG0160 (accession number according to http://www.oralgen.lanl.gov). The alignment of this ORF with two characterised
NADH oxidases is presented in Figure 4.16. As can be seen from that figure, the
ORF encodes a putative protein composed of 938 amino acids, while the two
NADH oxidases are shorter proteins, of about 450 amino acids. However, there are a number of important identities and similarities between the central region of
PGQ160 and the two NADH oxidases. As can be seen from the boxes in the figure, the cysteine in position 42 in the S. faecalis sequence is conserved in the other two sequences. This cysteine residue is an important feature of NADH oxidase, known to exist as a stabilised cysteine-sulphenic acid (cys-SOH) that serves as a non- flavin redox center (Ross and Claiborne, 1992). The other boxes indicate other partially homologous regions containing previously identified sequence fingerprints for FAD-containing, pyridine nucleotide dependent, oxidoreductases
(Ross and Claiborne, 1992). Boxes I and 4 are involved in FAD-binding and Box
3 involves contacts with NADH. Box 1 is not well conserved in the P. gingivalis sequence, but boxes 3 and 4 display high levels of homology.
On the other hand, when P. gingivalls was searched against AhpF, an ORF with high homology was identified. The alignment of this ORF, accession number
PG0556, according to http://www.oralgen.lanl.gov, and three other AhpFs is shown in Figure 4.I1. The arrows indicate the presence, in the four sequences, of two cysteine residues that have been shown to be involved in the catalytic activity
120 of the enzyme against hydrogen peroxide or cumene hydroperoxide (Niimura er a1.,1995). Conserved regions for FAD and NADH binding are also indicated.
Moreover, an ORF encoding for the second component of this system, the small sub-unit AhpC, was identified immediately upstream of the AhpF gene, a genetic anangement seen in other microorganisms (Poole et aL.,2000).
t2t Figure 4.16. Sequence alignment of PG 0160 (putative NADH oxidase-Nox)
and typical Nox proteins from Streptococcus faecalis and. Streptococcus
mutans
PG0 1 60 LVREAKNLRÀRTKKI SLHI CRKHAPQSE DFRCVFLRQQVI DRDTEAMAKPYDFY IVWKLS S. faec-NOX S.mutans-Nox
PG0160 YNHVFT ] HPRHT I LELPPVYPDTALVPFFIMRFSFFDHMT SCEY I FHRKLFVCLHRKNKQ S. faec-NOX S.mutans-Nox
FAD feglon Cys-42 PG0160 KQKSNMRYVI TA.AA,R -LRRI DEKAE I I LFEKGQNI SYA{õþLPYYTGGVI KE S. faec-NOX VKS I LANH-PEAEVTVYERNDN I S FLSCGlALYVGGVVKN S.mutans-Nox ----MSKIVI r LDNYGSENEVwFDQNSN r s rr,dCþualwr cKQr sc : *: ::::..***: .**:. ::* :.
PG0160 REN].FVQT PEKFGRLLNLDVRVQSEVLS I DRSDKQIRVRE-ANGREYSEHYDKLLLS PGA S. faec-NOX AADLFYSN PEEL-ASLGATVKMEHNVEE INVDDKTVTAKNLQTGATETVSYDKLVMTTGS S . mutans-Nox PQGLFYADKE SL YME S PVTAI DYDAKRVTA- .** *.: -EAKGAKI. : :: * *: . * : -LVNGOEHVESYEKL* *:**:::.*: I LATGS NADH-contact region PG0160 LPFVPPT,PGV -DSPGVFTLRNVEDTDAIKS-YLD-THKVKRÀIVVEæË]- S. faec-NOX wpr r ppr pGr - DAENTLLcKNysoANVT TEKAKDAK----RVVVVþccyr S.mutans-Nox TP I LPPIKGAÄIKEGSRDFEATLKNLQFVKLYQNAEDVTNKLQDKSQNLNRIAVVFAGY I *::**: * .: L::: : . * **E:la
PGo 1 60 GIEMAENLHARG IAVNVI EMAPOVMAPV-DFSMAT TVHAHLOEKG I GLYLGKAVKS IEKR I S. faec-NOX G It- E LVEAFVE S GKQVTLVDGL DRI LNKY],DKP FT DVLEKELVDRGVN LALGENVQQFVAD S.mutans-Nox GVELAEAFKRLGKEV] L ] DVVDTCLAGYYDQDLSEMMRQNLEDHG IELAFGETVKAI EGD I * * . 4*..* ' ::: * :: ::. .* ::*: * :*: *: : FAD-binding region PG0160 --GEVLTAS],DSGEKI EAEL I LLS I GVRPNTKLAÀDAQLAI GPARG T RVNEYLD'TSDPDT S. faec-NOX EQGKVAKVI TPSQEF-EADMVIMCVGFRPNTELL-KDKVDML PNGA] EVNEYMbTSNPDf S . mutans-Nox - -GKVERIVTDKASH- DVDMVI LAVGFRPNTALG-NAKLKTFRNGAFLVDKKQþTS I PDV *:* :.::::...*.**** * . . *.. l,** **.
PGO I 60 YAI GD.lAÏEYPHPLTGKPWTNF],AGPANRQGRIVADNMHGOTLRSYEGAT GTAIAKI FDL I S. faec-NOX FAAGDqAVVHYNPSOTKNY I P-LATNAVRQGMLVGRNLTEQKLAYR-GTQGTSGLYLFGW S . mutans-Nox YA] GDqATVYDNAINDTNYIA-LASNALRSG IVAGHNAAGHKLE SL-GVQGSNGT S I FGL :i--:til* I : :. , : ** * *.* :.. * :.* *. *: :*.
PGO160 TVAATGLPAKALKREGLPYE SVTVQPNSHAGYYPNA-YPLTLKI TFHPE SGMLYGAQCVG S, faec-NOX KI GS TGVTKESAKLNGLDVEATVFEDNYRPE FMPTTE -KVLMELVYEKGTQRIVGGQLMS S . mutans-Nox NMVSTGLTQEKAKREGYNPEVTAFTDFQKAS FIEHDNYPVTLKTVYDKDSRLVLGAQMÀS .: :**:. : * * * ... :, : : ::: :. : : *.*
PG0160 ]EGVDKRI DS IAQI IKRKGG IADLMQTEOAYAPPFS SAKDPVAT,AGYVADNI T TGRMKPL S. faec-NOX KYD I TQSANTLSLAVQNKMTVE DLAI SDFFFQPHFDRPWNYLNLLAOAA- S . mutans-Nox KEDMSMGI HMFSLATQEKVT IERLALLDYFFLPHFNQPYNYMTKAALKA- .: . :: ::,* : * : : * *, . : : , *
PG0 1 60 HV.]REMKEVDPNRVTLT DARPRQAYEVEH I PGAI SMPVEE ]RARI GE I PHDKP IYVYCAVG S. faec-NOX S . mutans-Nox
PG0160 MRGYFASN I LRQCGFSNVRNL IGGYRLYST I TADYSSAGOPATAQLPAKDS PS SHTQVPE S. faec-NOX ----LENt'f-- S . mutans-Nox ----:'------
122 Figure 4.16 (continued)
PG0 1 60 VDACGMSCPGP I LKLKQS I DQIAVGEQLC I LATDPGFARDAOAV{CDTTGHNL I RQETIKG S. faec-NOX S.mutans-Nox
PG0 1 60 KYKVT IEKTACKEEGTCVNETPSKGKTFI LFS DDLDKALATFVLANGAÀAMGQPVT I FFT S, faec-NOX S . mutans-Nox
PGO 1 60 FhJGLNAT KKPHAVKAKKDI ülGKMFGMMLPKN SKGLGL SKMNMFGIGAKMMRMVMKEKHVD S. faec-Nox S , mutans-Nox
PG0 1 60 S LE SMRKOALENGVE F IACQMSMDVMG I NREE],LDEVS T GGVATYMNRAEEAN INLFI S. faec-NOX S.mutans-Nox
123 ßig.4.17 Sequence alignment of PG 0556 (putative alkyl-hydroperoxide
reductase subunit F-AhpF) and typical AhpF proteins from Escherichia
coli, Streptococcus mutans and Amphibaccilus xyllanus.
PGO556 -MLDKDTLAQVGSYFAQLKKSYTLRINAHTSHPSYNEAKEMLDGLASVS DHVR-AEYNAA E. col-AhpF -MLDTNMKTQLKAYIEKLTKPVEL IATLDDSAKS -AE IKELLAE TAEI,S DKVTFKEDNSL S.mut-Nox/AhpF MALDAE I KEOLGQYLQLLECE IVLQAQI,KDDANS -QKVKEFLQE IVAMS PMI SL- DEKEL A. Xil-AhpF -MLDKDI KQQLEQYLALLENDIVIKVSVGDDKVS -KDTLELVNE IADMS SMI SV-EETTL ** : *: *. * : . *:: :. :* : :
PGO556 DDFRI DI,LVD- -GA-DSGI GFRG T PGGHEES SLLLAI LNNDGI GRNI P DEGVQDRIRRIN E. col-AhpF PVRKPSFL I TNPGS -NQGPRFAGS PLGHEFT SLVLALLWTGGHPSKE -AQSLLEQIRHI D S.mut-Nox/AhpF P-RTPS FRIAKKGQ-ESGVEFAGLPLGHEFYFVYLGSVTGFRASAKV-ETDIVKRI QAVD A. Xil-AhpF E-RT PS FS INRKGE PDSGVVFAGI PLGHEFTSLVLALLQVSGRAPKV-EASVI DAIKKVE .: : * :,* * * * **** : *. : : .: . *: ::
PGO556 GP]ELKTYV DVVQTLNMIAf LNPT INHTMVDGS FFPDEVE SLGIASVPTVMA E. col-AhpF GDFEFETYY SL DVVQALNLMSVLNPRI KHTAI DGGT FQNE I T DRNVMGVPAVFV S.mut-Nox/AhpF EPMHFETYV SL DVVQAPNIMSVVNPN I SHTMVEGGMFKDE IEAKG IMSVPTVYK A. XiI-AhpF GKHDFVTYV VVNPN I THTMVEGSAFKDEVDRLN I LAVP SVYL . ** ****..*,....** * ** '.* * .*
PG0556 GDEVI HVGRGDMAALLNKf EAKYGSVPAESADKTLRP IYSARKG E. col-AhpF NGKEFGQGRMTLTE IVAKI DTGAEKRAAEELNKR- DAYDVL I IYSARKG S . mut-Nox/AhpF DGTEFT SGRAS I EQLLDL IAGPLK- - -EDAFDDK-GVFDVLVI IYAARKG A. Xil-AhpF NGE FMASGRMTT EDI LGHLGSG I D- - -ASELNEK- DPFDVLVI SAIYAARKG . ** .*'*. ' . ***. ****
PGO556 LKVAIVAERVGGQVNETVGIENLI SVPYTTGSELASNLNSHIKANT I SLFEARTVS S ] T- E, col-AhpF IRTGLMGERFGGQI LDTVDIENYI SVPKTEGOKLAGALKVHVDEYDVDVI DSQSASKLT P IENMI GTPYVEGPQLMAQVEEHTKSYSVDIMKAPRAKS T - - S.nut-Nox/AhpF VKTGLLAETMGGOVMETVG _- A. Xi1-AhpF I RVG I IADRFGGQVMDTLGT ENEI GTKYTEGPKLT SE IEOHVNEYNI DVMKGLRAHN I :: .:: : .***' :*:.*** *.. . * :* . :: * :.::.. . .: Cys337 Cys340 PGo556 - -QQEG] SRVEVTSGEVFTAPAL IMÀTGASVÙRKLGVPGEKEYTGNGVA PFFKG E. co1-AhpF AAVEGGLHOIETASGAVLKARS T IVATGAKWRNMNVPGE DQYRT PLFKG S . mut-Nox/AhpF - - OKT DLVEVE LDNGAH LKAKTAVLALGAKh]RK I I LATGARWRD I GVPGEKEYKNKGVA A. XiI-AhpF - - EKKDLFEVQLDNGAVLKSKT.: .:: :.: **** .. .**.* NADH-binding region PG0556 KR I DLAGI CEHVTWEELDVLRADEVLQKKARETAN T DI LLSTATK E. col-AhpF KRVAV DLAG IVEHVTLLE FAPEMKADQVLQDKLRS LKNVDI I LNAQTT S.mut-Nox/AhpF KKVAV DLAGLASHVYI LEFLPELKADK] LQDRAEALDNI T I LTNVATK LNAQTT A. Xi1-AhpF SALDLAGIVKHVTVFEFMSELKADAVLQERLRSLPNVDVI*.****. ** ' **
PG0 55 6 E IMGNGOKVEGI ],LT DRNTGEEKQIALSGVFVQI GLAANT SLVKDL-VETNSRGEVL I DT E. cof-AhpF EVKGDGSKVVGLEYRDRVSGDI HNIELAGI FVQI GLLPNTNVùLEGA-VERNRMGE I I I DA S . mut-Nox/AhpF E f I GN- DHVEGLRYS DRTTNEEYLLDLEGVEVQI GLVPSTDWLKDSGLALNEKGE I IVAK A.xi1-AhpF E I TGD-ETVKGI SYI DRTTNEEKHVELQGAFIQI GLAPNTEWLGDT -VERNOIGE IVI DK *. *, * *. ** . . ' * * *.**** * . ' * **" '
PG0556 SCRTNTPGT YAAGDCTTVPYKQIVIAMGEGAKAÀLSAFEDRIRG- -- E. col-AhpF KCETNVKGVFAAGDCTTVPYKQI I IATGEGAKASLSAFDYL ]RTKTA S.mut-Nox/AhpF DGATNI PAT FAAGDCTDSAYKQI I I SMGSGATAÀLGAFDYLÏRN--- A.XiI-AhpF KGQTSVPG] FAAGDCT DTPYKQI IVSMGAGATAALGAFDYLLRN- -- * **.*:*.**: . 1. ..'****** .****..' :*
124 4.5 Expression of Nox and AhpF-C under anaerobic and oxygenated environments
To investigate if the putative NADH oxidase (nox) and alkyl-hydroperoxide reductase (ahpF and ahpC) genes were functional in P. gingivalis, the RNA expression levels for both of them, from cells grown under anaerobic and oxygenated environments, were analysed by northern blot hybridization.
4.5.1Methods
4.5.1.1 Growth of P. gingiv¿lls W50
P. gingivalis W50 was grown in continuous culture under the same growth conditions as described in section2.2.L2, in BM medium supplemented with 5 mg
L-l of haemin, to achieve haemin-excess, and 0.5 g L-l of cysteine. Total RNA was isolated at steady state from cells growing under the anaerobic atmosphere Nz/COu
(90:5) and under the oxygenated environment Nz/COzlOz (75:5:10).
4.5.1.2 RNA isolation
Total RNA was isolated by a modification of the method described by
Sambrook et al. (1989). Thirty mL of P. gingivalls cells were centrifuged at 6000 x
9,4"C for 2O min. The pellet was lysed in 1.5 mL ASE (20 mM sodium acetate pH
4.5, O.5Vo SDS, lmM EDTA). Phenol (1.5 mL) equilibrated in sodium acetate, pH
4.5, was then added to the lysed preparation. Samples were vortexed for 5 sec, incubated at 65"C for 5 min with gentle agitation and cooled for 5 min on ice.
125 Chloroform (0.3 mL) was then added and the samples were again mixed by gentle vortexing and centrifuged at 3000 x g to separate the phases. The upper aqueous phase was collected, phenol was added and the procedure described above repeated. The RNA was precipitated by incubation at -'70"C for 20 min in 3 volumes of ethanol followed by centrifugation (35000 x g, 4"C for 20 min). The collected RNA was dissolved in 0.5 mL of diethylpyrocarbonate (DEPC)-treated autoclaved distilled water and stored at -'70"C until further analysed. Prior to utilisation, all solutions used for RNA preparation and analysis were treated with
DEPC at a final concentration of O.IVo (v/v) and autoclaved in order to inhibit
RNAse activity.
4.5.1.3 Determination of RNA purity and concetttration
The RNA concentration of samples was determined by measuring the
ODzoonn'. Total RNA purity was checked by its ODzoo/OD2s6 ratio and by analysis in a non-denaturing ITo agarose gel (Sambrook et al., 1989).
4.5.1.4 Electrophoresis and blotting of RNA
Two pg of total RNA from each sample were separated by electrophoresis in a IVo agarose gel containing2.2 M formaldehyde and 1 x MOPS buffer (20 mM
3[N-morpholino] propane-sulphonic acid, 5 mM sodium acetate, 1 mM EDTA, pH
7.0). Prior to loading the gel, RNA (final volume 6 pL) was denatured at 65"C for
5 min in a solution containing I2.5 pL deionised fornamide,2.5 ¡tL 10 x MOPS buffer and 4 pL formaldehyde (37Vo). Samples were then chilled on ice for 5 min
and a solution containing 5OVo (v/v) glycerol with 0.1 mg ml--l bromophenol blue,
to a final concentration of IOVo, was added to each one. Gel wells were flushed
126 with running buffer (1 x MOPS) and run for 5 min at 70 V before the loading of samples; 3 pg of a 0.3-6.9 kb size RNA ladder (Roche) were used as a standard.
The gel was run for approximately 4-6 hours at 50 V or until the bromophenol tracking dye migrated at least 3/q of the gel length. It was then placed for 10 min in an ethidium bromide solution (5 pg ml--r;, de-stained in water for 30 min and photographed under UV light to confirm equal RNA loading.
RNA was then transferued to a Hybond N* nylon membrane by capillary elution with 20 x SSC buffer (0.15 M sodium chloride, 0.015 M sodium citrate, pH
7.0). Transfer was allowed to proceed for 16 hours after which the apparatus was dismantled and the membrane washed briefly in 2 x SSC buffer. RNA was fixed by exposing it to UV light for 3 min following which the membrane was air dried and stored at 4"C until hybridtzation and detection procedures were performed.
4. 5. 1. 5 Probe preparatiott
Primers were designed from the P. gingivalis W83 genome sequence database (http://www.tigr.org). Primers (left) CAGGTTATTGACCGGGACAT and (right) CGCTTTCGCCAGATAGAGAC were used to amplify a 900 bp fragment in the central region of the putative nox. Primers (left)
AGGAATATCCCCGATGAAGG ANd (Tight) CCGTTTCACGTGCTTTCTTT were used to amplify a 900 bp fragment of ahpF.
P. gingivalis genomic DNA was extracted according to the method of Chen and Kuo (1993). Briefly, 1.5 mL of a culture was harvested by centrifugation for 3 min at 10,000 x g. The cell pellet was re-suspended and lysed in 200 pL of lysis buffer (40 mM Tris-acetate pH 7.8, 20 mM sodium acetate, I mM EDTA, l%
SDS) by vigorous pipetting. To remove most proteins and cell debris, 66 pL of 5
121 M NaCl solution were added and the viscous mixture was then centrifuged (10,000 x g for 3 min at 4'C). After transferring the clear supernatant into a new vial, an equal volume of chloroform was added and the tube was inverted gently several times until a milky solution formed. Following centrifugation at 10,000 x g for 3 min, the extracted supernatant was transferred to another vial and the DNA was precipitated with IO0%o ethanol, washed twice with lOVo ethanol, dried and re- dissolved in distilled water.
This DNA was used as a template in a 50 pL PCR reaction containing 0.2 pM of each primer, I x PCR Gold Buffer, 1 mM MgCl2, 200pM each dNTP, 1.25 units AmpliTaq Gold DNA polymerase (Applied Biosystems). A PCR cycle consisted of a first denaturation step of 30 sec at 95"C, annealing at 55"C for 30 sec and extension af J2" for 1 min. 25 PCR cycles were allowed to proceed. An aliquot of the PCR product was analysed in a IVo agarose gel stained with 0.5 pg mLl ethidium bromide. The PCR product was then cleaned with a QlAquick@
PCR Purification Kit (Qiagen) following the manufacturer's instructions. The PCR product was then ligated into pGEMt-T Easy (Promega). Ligation was allowed to proceed overnight at 4"C, following which E. coli JM 109 High Efficiency
Competent cells (Promega) were transformed with the ligated product,
Transformants were selected on Luria Bertania plates supplemented with 0.5 mM
IPTG, 80 pg ml--r X-gal and 100 ¡rg mLt ampicillin. Plasmid DNA was prepared following standard protocols (Sambrook et al., 1989) and recombinants were identified, after digestion, with EcoRI (Promega).
For probe preparation, plasmid DNA was isolated, EcoRl-digested and the resulting 900 bp fragments were purified from a 7Vo agarose gel using the
128 UltraCleanrM GelSpin DNA Purification Kit (Mo Bio). DNA was then labeled using the DIG High Prime Labeling Kit (Roche) and used for hybridizafion.
4.5.1.6 Northern hybridization and detection
The membrane was pre-incubated for 30 min at 42"C in DIG Easy Hyb buffer (Roche), placed in the DNA ahpF probe solution (final concentration 25 ng ml--r in DIG Easy Hyb buffer) and hybridized overnight at the same temperature.
The membrane was then washed twice in 2 x SSC, 0.1% SDS at 2O-25"C for 5 min and twice in 0.5 x SSC,0.1% SDS at42"C for 15 min under constant agitation.
Detection of the hybridized DNA was performed using the DIG High Prime DNA
Detection Kit (Roche) utilising the chemiluminescent substrate CSPD and exposure for 30 min to X-ray film.
For the nox probe, different concentrations of probe (25 and 100 ng mL-t; and different hybridization temperatures (31 , 42 and 48"C) were utilised.
4.5.2 Results
4.5.2.1 Expression of AhpF under anaerobic and orygenated environments
Northern analysis of P. gingivalis total RNA from the two conditions, anaerobic and oxygenated growth, with a DlG-labeled ahpF-specific probe is shown in Figure 4.18. A transcript of approximately 2.2 kb appears in cells growing anaerobically as well as under oxygen. The size of this transcript indicates
that ahpF is transcribed in a polycystronic way with its upstream neighbouring
gene ahpC. AhpF is 1545 bp, ahpC is 564 bp and there is an intergenic region between the two of 161 bp, for a total transcript size of 2276 bp. The transcription
129 of this fragment appears to increase slightly under oxygenated conditions, as judged by the intensity of the bands.
Figure 4.18. Northern blot analysis of AhpF mRNA in P. gingivølis steady
state cells grown under anaerobic (Nz) and oxygenated (Oz) environments.
1 2 kb N2 02 N2 02 kb
235 2.1 2.7
1.8
1.5
l. Photograph of membrane, that was ethidium bromide stained to confirm equivalent sample loadings. 2. Northern blot of membrane in panel 1 following hybridization with an AhpF specific probe.
4.5.2.2 Expression of Nox under anaerobic and oxygenated envirotxments
Although different probe concentrations and different hybridization temperatures were utilised for the detection of mRNA corresponding to the putative NADH oxidase, no signal was detected on the membranes under any of the conditions tested.
130 4.6 Summary of results and discussion
In the first experiments described in this chapter, the activities of three enzymes, NADH oxidase, NADH peroxidase and SOD were assayed in P. gingivalis cell extracts. Activity of each of the three enzymes was detected under all the conditions tested, and was enhanced when growth occurred under oxygenated environments. In particular, the increase in activity of NADH oxidase and NADH peroxidase seemed to be greater in the presence of oxygen than the increase in SOD. As the role of SOD has been investigated extensively (Amano e/ al. , 1990; Lynch and Kuramitsu, 1 999), the main focus of this study was to clarify the role of NADH oxidase in the physiology and oxidative stress response of P. gingivalis.
The NADH oxidase activity in cell extracts presented the typical features seen in other NADH oxidases; it was stimulated by the presence of FAD and p-
NADH was the preferred electron donor.
Subsequently, in order to identify the protein responsible for NADH oxidase activity, a purification of the activity was carried out. One obstacle found during the NADH oxidase isolation was that the addition of the proteinase inhibitor
TLCK decreased the NADH oxidase activity by more than half. The mechanism by which TLCK inhibits the cysteine proteinases is thought to be through the formation of a covalent bond with a histidine residue in the active site. In this respect, TLCK might not be specific for the proteinases, as histidine residues might also play a role in other enzymes. Therefore, the low yield obtained during the purification procedures might have been a reflection not only of the low NADH oxidase activity in P. gingivalis cell extracts, compared to other microorgansims
(see Section 7), but possibly of proteolytic cleavage during purification.
131 Nevertheless, despite the low yield, it was possible to partially purify the protein with NADH oxidase activity. During purification of the NADH oxidase active protein, activity was not only monitored by cuvette assay, but also by assays in non-denaturing gels (as a control). These gels showed that a single protein had significant NADH oxidase activity in P. gittgivalis cell extracts during all the purification steps and were particularly useful after ammonium sulphate fractionation, when different fractions were found to possess NADH oxidase activity. The gel assay indicated that the same protein was responsible for the activity in all the fractions and allowed confident continuation of the purification of the fraction presenting with the highest NADH oxidase activity.
After purification, it was found that the NADH oxidase activity was associated with 4-hydroxybutyryl-CoA dehydratase, an enzyme involved in the metabolism of glutamic acid. Although this finding was sutprising, it was confirmed by assaying the NADH oxidase activity of preparations of the same enzyme from another microorganism, C. aminobutyricum. However, as the NADH oxidase activity of this enzyme was 1000 times lower than the activity towards its natural substrate, 4-hydroxybutyryl-CoA (Dr. Wolfang Buckel, personal communication), the biological significance of the NADH oxidase activity displayed by the dehydratase is unclear.
The mechanism by which 4-hydroxybutyryl-CoA dehydratase is capable of showing NADH oxidase activity is also unclear. The dehydratase is known to bind
FAD and so if it also reacts with NADH, it could reduce the FAD to FADHz; which would, in turn, react with oxygen to produce HzOz. This last reaction could
occur spontaneously and not necessarily be enzyme-mediated. However, this
732 activity would be contrary to the main activity of the dehydratase - on 4- hydroxybutyryl-CoA and requiring the oxidised FAD.
The activity of C. aminobutyricum AbfD towards its main substrate, 4- hydroxybutyryl-CoA, is inactivated by oxygen (Müh et al., 1991). If a similar inactivation occurs with P. gingivalis AbfD, then this could explain why the production of butyrate is decreased in oxygen stressed cells. However, 4- hydroxybutyryl-CoA dehydratase also catalyses the isomerisation of vinylacetyl-
CoA to crotonyl-CoA. This second activity is only partially inactivated by oxygen.
Hence the oxidase activity could be resistant to oxygen and still be present in oxygen stressed cells. However, as it is not likely that the cells induce the production of more dehydratase under oxygenated conditions, the increase in the activity of NADH oxidase in response to oxygen, might signify that the dehydratase is not the sole enzyme with NADH oxidase activity measured in cell extracts.
It is possible that anothel eîzyme, also with NADH oxidase activity, could be induced under oxygenated conditions and equate to the activity presented by the remaining dehydratase. This enzyme is most likely to be alkylhydroperoxide reductase (AhpF-C), which is also capable of NADH oxidase activity (Niimura er al.,2OOO).Indeed, an increase in the mRNA for AhpF-C was seen to occur in P. gingivalis under oxygenated conditions. The degree to which AhpF-C contributes to the overall NADH oxidation measured in P. gingivalis cell extracts is unknown; but it might be minimal compared to the NADH oxidase activity displayed by
AbfD, as no additional bands with activity were detected in non-denaturing gels of cell extracts when assayed for NADH oxidase. Alternatively, the AhpF may not have dissolved in the gel buffer, and as a consequence did not penetrate the gel.
133 Holvever, this is unlikely, as only proteins with very high pls usually behave in this manner. (The AhpF theoretical pI is 4.9).
The sequence alignment of AhpF of P. girtgivalis and that of A. xylanus showed high degree of homology. Therefore, it is likely that this protein functions in a similar manner in P. gingivalis. AhpF could work as an H2O2-producing
NADH oxidase under aerated conditions in the absence of HzOz. AbfD might also generate HzOz through a similar reaction. The HzOz generated from both proteins could then be metabolized by the AhpC component.
The proposed reaction mechanism for NADH oxidase activity of A. xylanus
AhpF-subunit is as follows:
Figure 4.19 NADH oxidase mechanism of A. xylanzs AhpF protein (Niimura
et ø1.,1995).
FAD S I F S NADH Hzoz +H+ A. rylanus NADH oxidase
(AhpF) O2 NAD+ FADH2 s I F S + I FAD SH F SH
NADH + H+ + Oz NAD+ +H2O2
-
134 The hydrogen peroxide produced by the AhpF could then be removed by the AhpC component, which is reduced by the electrons that pass from FADHz through Cys-337 and Cys-340 in the AhpF sub-unit to the disulphide center of
AhpC (Niimura et a|.,2000).
Therefore, when an NADH oxidase assay is performed aerobically on P. gingivalis cell extracts, the reduction of NADH might be the result of the reactions seen in Figure 4.20:
Figure 4.20. Proposed mechanism for the conversion of IJzOz to HzO
by A. xylanøs AhpF-C system (Niimura et a1.,1995).'
Hzot
SH AhpC soH FAD t SH F SH S 2 NADH I +2H+ F s 2 l--s AhpF I¡ Hzo
FADH2 FAD AbfD? SH SH F SH F SH
2 NADH + 2H+ * 02 2 --+ NAD+ + 2H2O u The mechanism by which AbfD generates HzOz via the oxidation of NADH is unknown. P. gingivalls AhpF-C might function similarly to A. Xylaturs AhpF-C
135 Thus, these Figures explain why the addition of catalase did not decrease the rate of oxidation of NADH, although HzOz was being produced (and consumed) by the cell extracts. As shown in Fig. 4.20, the metabolism of H2O2 by
AhpC does not depend on NADH; and so, in the presence of catalase, the deprotonation of NADH will continue through AhpF and AbfD, while catalase will immediately convert the HzOz produced to water and 02, suppressing only the non-
NADH-dependent AhpC activity.
On the other hand, when H2O2 is present, and in the absence of oxygen, as when NADH peroxidase activity was assayed in P. gingivalis cell extracts, the following reaction could occur:
Figure A.2lReaction mechanism for hydrogen peroxide reduction catalysed
by A.xylanus AhpF-C (Niimura et a1.,1995).
IJzOz HzO
FSH SOH FAD l- tt F SH S AhpC NADH I +H+ F S NAD+ Fs AhpF l-$ Hzo
FADH2 FAD S SH I F S F SH
NADH + H* + HzOz -+ NAD + 2HzO
136 Moreover, it is likely that NADH would still be able to reduce the dehydratase under anaerobic conditions and this reaction would still contribute to the NADH deprotonation measured in the NADH peroxidase assay.
It is worth noting that the stoichiometry and precise mechanisms of the leactions described have not been established in P. gingivalis. Thus, the proposed mechanisms are based on the assumption that the P. girtgivalis AhpF-C system works in a similar fashion to that present in A. xylanas (to which P. gingivalis
AhpF-C shows a high degree of homology).
The mechanism by which P. gingivalis AbfD would operate as an NADH oxidase might possibly be related to some cysteine residues in the protein capable of forming disulphide bonds (as occurs in other NADH oxidases). In this respect it is interesting to note that in the alignment of P. gingivalis AbfD with C. aminobutyricum and C. dfficile AbfDs (Fig. a.10), three cysteine residues, not involved in the formation of the [4Fe-4S] cluster are conserved among the three sequences. These cysteines, therefore, could act as the redox center for the NADH oxidase activity displayed by the dehydratase.
The presence of an NADH oxidase of the type of S. mutans H2O-producing
Nox, was also investigated in P. gingivalis genome sequence. The results presented in Sections 4.4 and 4.5 show that although an ORF with homology to a Nox protein was identified, no signal could be detected in the northern blot analysis, despite the different hybridization and washing conditions tried. Failure to detect a signal in the northern blot, although not conclusive, might suggest that either no
RNA molecule for this protein is transcribed or that the transcribed RNA molecule is not stable. In this respect, it is interesting to note that the ORF encoding for the putative Nox, identified in the genome sequence, does not coffespond to the size of
131 the oxidase. An analysis of the region upstream of the putative NADH oxidase
Fig. a.22) shows what appears to be a promoter sequence with similarity to a consensus for different P. gingivalis genes reported by Jackson et al. (2000), with the putative transcription start point located at nt -55 (55 nucleotides upstream of the translation start codon methionine (+1)). This type of arrangement, with translation start-sites located some distance from the transcription start point, is a common feature in P. gingivalis and other members of the Bacteroidaceae
(Jackson et al., 2000). However, the last residue (Ala) in the putative protein is located at nucleotide 1708 of the ORF (composed of 2814 nucleotides) and no stop codons or a transcription termination site are found downstream. Therefore, it seems that although the arrangement of nucleotides might favor transcription, the
RNA molecule would be at least 1000 nucleotides longer than the size of the RNA required for translation into the NADH oxidase.
In summary, the experiments presented in this chapter suggest that the majority of the NADH oxidase activity measured in P. gingivalis cell extracts, might be due to an NADH oxidase-like activity displayed by 4-hydroxybutyryl-
CoA dehydratase. However, the biological significance of this activity is not clear as it seems that the activity of the dehydratase for its natural substrate, 4- hydoxybutytyl-CoA, would compete with its NADH oxidase activity since both require the involvement of FAD. An analysis of the expression of other proteins likely to function as NADH oxidases/peroxidases in P. gingivalis revealed that transcription occurs for a protein similar to alkyl-hydroperoxide reductase, which could have a role as an NADH oxidase-peroxidase system. The transcription of the mRNA specific for this protein is increased under oxygenated conditions, compared to anaerobic growth.
138 Figure A.22.Seqluence analysis of ORF PG 0160 that encodes for a putative
Nox (indicated in amino acid codes-upper case). Possible promoter sequences are
in bold and underlined, consensus upstream sequence is in bold. Possible transcription
start point is in upper case. No stop codons or transcription termination sequences were
found downstream of the Nox sequence. x indicate nucleotides identical to the consensus
promoter sequence reported by Jackson et al., (2000).
ttggttcgggaa5caaaaaatttacgcgcgagaacgaaaaaaatctcgctccacattfgcaggaaacacgcaccgcaatcggagg attttcggtgcgtatftct5agacagcaggttattgaccgggacatcgaagcaafggccaaaccttatgaÍttctatatcgtttggaaact g|ccLacaatcacgtctttaccatacafccaaggcataccatcttagagctgccacctgtctatcccgacaccgctctcgttccgttcttc * attatgc gtttttcatttttc ggaaaaacaaa nt caaaaacagaaatc aaaTMRYVIVccVAGGATAAARLRRIDEKAEIILFEKGQNIS YANCGL
PYYIGGVIKERENLFVQTPEKFGRLLNLDVRVQS EVLS IDRS D KQIRVREANGREYS
EHYDKLLLS PGALPFVPPLPGVDS PGVFTLRNVEDTDAIKS YLDTHKVKRATVVGG
GFIGLEMAENLHARGIAVNVIEMAPQVMAPVDFS MATTVHAHLQEKGIGLYLGKA
V KS IEKRGEVLTAS LD S GEKIEAELILLS IGVRPNTKLAADAQLAIGPARGIRVNEYL
QTS DPDIYAIGDAIEYPHPLTGKPWTNFLAGPANRQGRTVADNMHGQTLRS YEGAI
GTAIAKIFDLTVAATGLPAKALKREGLPYES VTV QPNS HAGYYPNAYPLTLKITFHP
ES GMLYGAQCVGIEGVDKRIDSIAQIIKRKGGIADLMQTEQAYAPPFS SAKDPVAL
AGYVAgacaatatcattacaggtcgtatgaaacctctgcattggcgcgaaatgaaagaggtcgaLcccaafcgagtaaccctta tagatgctcgtcccagacaggcctacgaagtagagcatatcccgggagcaafcagcatgccggtggaggaaataagggcacEca laggggaaattcctcacgacaaaccgafctatgtafattgcgccgtggggatgagggggtatttcgcctcgaacatcctgagacaat gtggtttctcaaatgfgcgaaacctgatcggtggctatcgcttgtaftctaccataacggcagactacagttctgccggccaacctgcc accgcccaacttcctgcaaaggatagtccttcttcccatacacaggtaccggaggtggatgcctgtggcatgtcctgtccggggcc gafcctcaagctcaaacaatccatfgaccaaatagclgtaggagagcagctctgcatc cfggctaccgaccccggatttgctcgtga f5ctcaagcctggtgcgacacgaccggacacaatctgatccggcaggagaccataaagggcaaa|acaaagtgaccatcgaaaa
139 F.ig.4.22 (Continued)
gactgcttgcaaggaggaaggcacctgcgtgaacgagactccatccaaaggcaagacattcattctcttcagcgatgatctggaca aggcttfggccacattcgtattggccaacggagcagcagcgatgggacagccggtgactatatfctttactttctgggggttgaatgc catcaaaaagcctcacgccgtcaaagcgaagaaagacatttggggcaagatgtfcggtatgal9ctfccgaagaacagcaaaggg ttaggactgtcc aaaatgaacatgttcggcttgggagcgaagatgatgcgtatggtcalgaaggaaaagcatgtggactcactgga gagcatgcgtaaacaagcgttggaaaacggtgtagagtttatagcctgtcagatgtccatggatgfaatggggatcaatcgggaag aatt5ctcgacgaggtaagtafcggtggagttgccacctatatgaatagagcagaagaggccàacafcaatclattcatc
140 5. Studies on possible regulatory systems of the
oxidative stress response in P. gingivalis
141 5.1 Identification of possible transcriptional regulators of the oxidative stress response in the P. gingívøfis genome sequence
5.L.1Methods
Sequences encoding for transcriptional regulators such as OxyR, SoxRS,
OhrR, PerR and OxyS which are involved in the control of the response to different kinds of oxidative stress [for a review see Storz andZheng (2000)], were obtained from the NCBI Entrez database (http://www.ncbi.nlm.nih.gov) (Table
5.1). These sequences were searched, one by one, against the P. gingivalis nucleotide database translated into protein, using Blastp (Gish, 1996-2000) at The
Institute for Genomic Research Comprehensive Microbial Resource (tigr cmr)
(http://tigrblast.tigr.org). OxyR, the only transcriptional regulator identified, was aligned with other OxyR proteins from different microorganisms (Table 5.1) using the program T-coffee (http://www.ch.embnet.org/software/tcoffee.html).
142 Table 5.1. Sequences obtained from NCBI Entrez database used to identify transcriptional regulators of the oxidative stress
response in P. gingivalis, or used in the alignment with PG 0242, putative OxyR.
Protein or gene name Microorganism Accesion number Reference
SoxS-SoxR S almone lla ty phimurium U6II47 (Martins and Demple, 1996)
SoxS-SoxR Escherichia coli M60111 (Wu and Weiss, 1991)
OxyR Yersinia pestis NP 407360 (Parkhill et aL.,2001)
OxyR Escherichia coli x16531 (Tao et a\.,1989)
OxyR Stre ptomy c e s c o e lic olo r AAF0674s (Hahn and Roe, 1999)
OxyR Salmonella enterica NP 458s68 (Parkhill et a|.,2001)
OxyR H aemo phil u s infL uenlae u49355 (Maciver and Hansen, 1996)
PerR Escherichia coli u57080 (Loewen, 1996)
OxyS Escherichia coli u87390 (Altuvia et aL.,1997)
OhrR Xanthomonas c amp e s tri s AF036166 (Mongkolsuk et al., 1998)
143 5.1.2 Results and discussion
After a Blast search against the P. gingivalis genome sequence, the only sequence of an oxidative stress related regulon producing high scoring segment pairs was OxyR. An alignment of PG 0242 (putative OxyR) and other OxyR proteins from Haemophilus influenzae, Escherichia coli, Streplomyces coelicolor andYersinia pestis is shown in Figure 5.1.
Although some of the above organisms are not closely related, the consensus alignment shows homologies that extend across the protein sequences.
In particular, the fact that cysteine residues in positions 199 and 208 in most of the sequences, are conserved, is an indication that all of these OxyR proteins might be functional. These cysteine residues have been shown to be critical for the ability of the transcription factor to sense HzOz itt vivo and in vitro (Zheng et al., 1998).
Moreover, the helix-turn-helix motif region for DNA binding and promoter recognition, present at the N-terminal domain of LysR-type regulators (Kullik ør al., 1995), is also highly conserved among the OxyR sequences, especially if the
Streptomyces coelicolor sequence is disregarded.
144 Figure 5.1 Sequence comparison of PG 0242 and other OxyR
homologues from different organismsu.
LysR family helix-turn-helix motif
PG0242 ------MNTQQ AKAADYCNVSQPTLSTMI LKLEEE DRSRQ H. infl -----_-MNIRDLEYL SCNVSQPTLSGQIRKLEDELGI RTSR E. coli ------MNI RDLEYLVALAE SCHVSOPTLSGQIRKLEDE S , coel- MS SKKRQPSLAQLRAS DAVAE GMSQPALSGAVSALEEALGLT Y. pest - -MNI RDLEYLVALAE ù IRKLEDELG] - - - - - . .* *,
Pc9242 P IE PTS I GALVVSQAKQI LYDLNS I TRI IEEEQOSLTGRLNIAVLPTTAPYLLPRVFP Ï!Ù H. infl KVLFTQSGMLLVDQARTVLREVKLLKEMASNOGKEMTGPLH I GLI PTVGPYLLPY IVPML E. coli KVLFTQAGMLLVDQARTVLREVKVLKEMASQQGETMSGPLHI GLT PTVGPYLLPHI I PML S. coel KVLL S PAGARLAVRTKAVLAEVGALVEEAEAVRÄP FTGALRLGV I PTVAPYVLPTVLRLV Y. pest KVLFTQAGLLLVEQAKTVLREVKVLKEMÀSLQGE SMSGPLHT GLI PTVGPYLLPOI I PML * . .* ,. .* * . ..**. **.** .
PGj242 KKELAGLE IHVSEMQT SRCLASLLSGE I DMAI IASKAETEGLEDD],LYYEE FLGYVSRCE H. infl KAAFPDLEVFLYEAQTHQLLEQLElGRLDCAIVATVPETEAFTEVP I FNEKMLLAVSEHH E. cofi HQT FPKLEMYLHEAQTHQLLAQLDSGKLDCVI LALVKE SEAFIEVPLFDE PMLLAI YEDH S. coel HERYPDLDLQVHEEQTAS LLEGLTTGRLDLLLLAVPLGVPGVTELPLFDE DFVLVT PLDH Y. pest HKTFPKLEMYLHEAQTQNLLAQLDSGKLDCAI LALVKETEAFIE ] PLFDE PMNLAÏYADH *.. . * ** .* .* ..* cvs-iq's 'Cys-208 PG0242 PLFEQDVIRTTEVNP LH- -EROTAYS GGSMEAFMRL H. infl P!ÙAQESKLPMNQLNG --E_NSHFOATSLETLRNM E. coli P!{ANRECVPMADLAGEKLLMLE -E-DTHFRATSLETLRNM S. coeÌ RLGGREGLERSVLRELKLLLL P-ATTTTAÄ,GLSTLVQL Y. pest PWANRERVEMHELAGEKLLMLE, .*. -E- DTHFRATSLETLRNM
PGj242 VESGQG I TFI PQLTVEQLS PSQKELVRPFGM- PRPVREVRLAVRQDYSRRKLREQL Ï GLL H. infl VAANAG T TFMPELAVLNEGTRKGVKYI PCYS - PE PSRT]ALVYRPGS PLRNRYERVASAV E. coÌi VAAGSGI TLLPALAVPPERKRDGVVYLPC IK- PEPRRT IGLVYRPGS PLRSRYEQLAEAI S . coel- VAGGLGVTLLPRTAVRVETSRS SQLLTAS FTDPAPTRRIALAMRTGAARSAEYGELAAAL Y. pest VAAGSGI TLLPALAVPNERQRDGVCYLECYK- PVPKRT IALVYRPGS PLRGRYEQLAEAI * *.*..* .* * * * . * *
PG0242 RSAVPSDMHKLQTGQHLA H. infl SDEVKS ILDGLK------E. coÌi RARMDGHFDKVLKOAV-- S. coef REAMADLPVRTVHD---- Y. pest RDHMQERMAPSLEQAI --
u Critical cysteines and helix-turn-helix motif region for DNA binding and promoter recognition
are indicated.
145 5.2 Construction of a P. gingivalis OxyR-inactivated mutant
In order to study the role of OxyR in the P. girtgittalis oxidative stress response, a mutant was constructed by insertional inactivation of oxyR with a tetracycline resistance cassette.
5.2.1. Methods
5.2.1.1 Primers and oxyR amplification
The primers TCGAATACATAGCCGCATTG (forward) and
CCTGTCTGCAACTTGTGCAT (reverse) were designed from the ORF PG 0242 encoding a putative OxyR in P. gingivalis W83. A 902 bp fragment was amplified by PCR as follows: P. girtgivalis genomic DNA, prepared by the method of Chen and Kuo (1993) was used as a template in a 50 ¡rL PCR reaction containing 0.2 pM of each primer, 1X PCR Gold Buffer, 1 mM MgCl2, 200pM each dNTP and I.25 units AmpliTaq Gold DNA polymerase (Applied Biosystems). A PCR cycle consisted of a first denaturation step of 30 sec at 95oC, annealing at 55"C for 30 sec and extension at J2"C for 1 min. Twenty-five PCR cycles were allowed to proceed and then an aliquot of the PCR product was analyzed in a IVo agarose gel stained with 0.5 pg mLt ethidium bromide. The PCR product was then cleaned with a QlAquick@ PCR Purification Kit (Qiagen) following the manufacturer's instructions.
146 5 .2. 1 .2 Construction of an OxyR: : tetQ DNA fragntent
The oxyR PCR product was then ligated into pGEM@-T Easy (Promega) as previously described. Transformants were selected on Luria-Bertani plates supplemented with 0.5 mM IPTG, 80 pg mLr X-gal and 100 pg mLt ampicillin.
Plasmid DNA was prepared following standard protocols (Sambrook et a\.,1989) and recombinants were identified after digestion with EcoRI (Promega).
Plasmid pNJR12 in host E. coli JM 109, was obtained from Dr. Nada
Slakeski, The University of Melbourne. This plasmid contains a tetracycline resistance gene (tetQ) cloned from Bacteroides thetaiotamicron, confering tetracycline resistance to P. gingivalis and not to E. coli (Maley et al., 1992). E. coli JMI09 cells containing pNJRI2 were grown on LB medium supplemented with 50 pg mLt kanamycin. TetQ was removed from pNJR12 by digestion with
EcoCRI (Promega) and purified from a l%o agarose gel using the UltraCleanrM
GelSpin DNA Purification Kit (Mo Bio).
The pGEM@-T Easy vector containing the oxyR fragment was then digested in a unique site in oxyR with BsaML The DNA was then cleaned with the
QlAquick@ PCR Purification Kit (Qiagen) and the protruding 3' tails blunt-ended with T4 DNA polymerase (Promega). The DNA was cleaned again and subsequently dephosphorylated with Shrimp Alkaline Phosphatase (Promega). The previously-prepared tetQ insert was then ligated into this fragment and the resulting pGEM@-T Easy::o.r7R::tetQvector was used to transform E. coli JM 109.
Plasmid DNA was then EcoRl-digested to check for colonies containing the correct insert.
141 5.2. 1. 3 P. gingivalis transþnnatiort
P. gingivalls W50 cells were prepared as follows. A 200 mL overnight culture, grown in Brain Heart Infusion medium (Oxoid) supplemented with 5 pg ml--l of haemin and 0.5 g L-r of cysteine, was pelleted by centrifugation (6000 x g,
4oC for 15 min). Cells were then washed in an ice-cold sterile l}Vo glycerol solution containing 1 mM MgCl2, pelleted again and resuspended in 0.4 mL of the same solution. Eighty pL of cells were then mixed with approximately 200 ng of the linear oxyR::tetQ DNA fragment (gel purified) and electroporated at 1.8 kV.
Cells were then allowed a recovery period in 1 mL BHI for 2 h under an anaerobic atmosphere (Nz:COz:Hz; 90:5:5) and plated on blood agar containing 1 pg mL' of
Tetracycline. Colonies were usually recovered after 5-6 days of anaerobic growth at37"C.
5.2.1.4 Southent blot and PCR analysis of recombinant clones
P. gingivalis cells obtained from each tetracycline resistant colony and P. gingivalis wild-type cells were grown overnight in 20 mL BHI medium supplemented with 5 pg mL 1 of haemin and 0.5 g Lr of cysteine. Genomic DNA was then prepared according to the method of Chen and Kuo (1993). The protocol included the removal of RNA using an RNAse cocktail (Promega) in the lysis step for 30 min, at 37"C. Genomic DNA was analysed in a O.lVo agarose gel and digested with Bgl tr to produce a 306i bp fragment containinB oxyR or a 5.718 bp fragment containing oxyR::tetQ. DNA was electrophoresed in a 0.7Vo agarose gel at 60V for approximately 4 hours, stained with ethidium bromide and photographed. Subsequently, the gel was rinsed for 10 min in 0.25 M HCl, 40 min in denaturing solution (0.5 M NaOH, 1.5 M NaCl) and 2 x 20 min in neutralising
148 solution (0.5 M Tris HCl, pH 7.5, 1.5 M NaCl) following standard protocols
(Sambrook et a\.,1989). DNA was transferred to a Hybond N* nylon membrane by capillary elution with 20 x SSC buffer (0.15 M sodium chloride, 0.015 M sodium citrate, pH 7.0). Transfer was allowed to proceed for 16 hours, after which the apparatus was dismantled and the membrane washed briefly in 2 x SSC.
Subsequently, the membrane was hybridised against an cr-32p-dATP- labelled oryR DNA probe or against an a-32p-dATP-labelled tetQ DNA probe, previously prepared using the Prime-a-Gene labeling System (Promega).
Hybridization was allowed to proceed overnight at 68"C in hybridization solution containing 6 x SSC, 0.25Vo SDS, 3 x Denhardt's solution (Sambrook et al., 1989) and 100 pg mL' hening sperm DNA. The membranes were then washed thoroughly tn2x 20 min 2 x SSC,2x20 min 1 x SSC and2 x 20 min 0.5 x SSC at
68"C and exposed to X-ray film for probe detection.
Positive colonies were also confirmed by performing a PCR reaction using the primers TCGAATACATAGCCGCATTG (forward) and
CCTGTCTGCAACTTGTGCAT (reverse), used previously for the amplification of oxyR from P. gingivalis chromosomal DNA.
5.2.2 Results
The hybridization of the o.rryR-specific probe was expected to produce a
3.061 kb band in the wild-type strain and a 5.718 kb band in the OxyR- mutant. On the other hand, hybridization with the tetQ-specific probe to the wild-type DNA was expected to result in no band, while hybridization to the OxyR mutant was expected to result in a 5.118 Kb band. Figure 5.2 shows that all the expected results were obtained in the southern blot analysis.
t49 Figure 5.3 shows a lVo agarcse gel with the PCR products obtained from the wild{ype and the OxyR- strains, utilising primers that amplify most of oxyR.
The primers used produced a 902 bp product in the wild-type and a -3700 bp fragment in the mutant, which together with the southern blot, confirmed the correct tetQinsertion in o.r7R.
Figure 5.2. Southern blot analysis of Bglll-digested genomic DNA from P. gíngivalis W50 wild-type and OxyR'strains, utilising an o-32p-dATP-labelled
tetQDNÃ probe (1) and a ct-32p-dATP-labelled. oxyR DNA probe (2).
(1) (2)
wild- witd- OxyR- OxyR
23.1
9.4 6.6
4.4
2.3 2.0
150 Figure 5.3. PCR products of P. gingívalis W50 wild type and OxyR' strains.
v/ild- OxyR kb type
l0 8 6 5 4 -t 25 2 l5 I
08
06
o4
151 5.3 Characterisation of the P. gingivølis W50 OxyR- mutant
5.3.1Methods
5.3.1.1 Effect of oxyR disruption on anaerobic growth
To characterise and compare the growth of the OxyR-inactivated mutant with the wild-type, both strains were grown overnight under anaerobic conditions in 20 mL BHI medium supplemented with 5 pg ml--t of haemin and 0.5 g L-r of cysteine. Two mL of this culture were used to inoculate a 100 mL broth of the same medium that was incubated anaerobically at 3J"C and monitored every hour for its ODsoon,n until stationary phase was reached.
5.3.1.2 Air sensitivity on solid media of OxyR and wild-type
To compare the resistance of individual cells to killing by atmospheric oxygen, 0.1 mL of serial dilutions of wild-type and mutant cultures were spread on blood agar plates which were exposed to air for different time periods and then incubated anaerobically for 4-6 days, as described previously (Section 2.1). The number of colonies appearing on a plate after a specific time of exposure to air was compared to that obtained for the same dilution on a control plate not exposed to air. The percentage of c.f.u. appearing on plates exposed to air compared to the control (unexposed) was determined for the wild-type and for the OxyR- mutant.
5.3.1.3 HzOz sensitivity in liquid cultures
An initial experiment was performed to determine the minimum inhibitory concentration (MIC) of HzOz for both the OxyR- mutant and wild-type. Both
152 strains were grorwn to mid-logarithmic phase, under anaerobic conditions, in 20 mL
BHI supplemented with 5 pg ml--t of haemin and 0.5 g L-r of cysteine. These cultures (0.5 mL) were used to inoculate, in the anaerobic chamber, 4.5 mL of fresh broths of the same medium. Broths contained different HzOz concentrations ranging from 0 to 50 mM. For every H2O2 concentration tested, duplicate broths of each strain were inoculated. One of them was incubated in the anaerobic chamber
(Nz:COz:Hz; 90:5:5) at 37"C while the other was incubated, with the lid slightly
open, under normal atmospheric conditions at the same temperature. After 16 hours of growth the ODsoonm was determined.
After the MIC for HzOz was determined for each strain, a second experiment was carried out in which the effects of different HzOz concentrations
on the (batch) growth curves of P. girtgivalis wlId-type and OxyR strains were
determined. Both strains were grown overnight under anaerobic conditions tn 20 mL BHI supplemented with 5 ¡rg mLt of haemin and 0.5 g L-l of cysteine. Two mL of this culture were used to inoculate a 100 mL broth that was monitored for its
OD560n' every hour until stationary phase was reached. Hydrogen peroxide, to the
desired concentration, was added at early exponential phase (approximately 9 hours after inoculation).
5.3.2 Results
5.3.2.1 Effect of oxyR disruption on the anaerobic growth of P. gingivalis
The anaerobic growth curves of P. girtgivalisW50 wild-type and its OxyR mutant are compared in Fig. 5.4. The two strains exhibited the same growth rate
153 during logarithmic phase growth and grew to approximately the same cell densities at stationary phase.
Figure 5.4. Anaerobic growth of P. gingívalis W50 wild-type and OxyR-
mutant.
1.8 1.6 1.4 1.2 +Wild-type ODsoon- 1 -+-Ox¡¡R- 0.8 0.6 0.4 0.2 0 0 10 20 30 Time
5.3.2.2 Air sensitivity of the OryR mutant
Table 5.2 shows the ability to survive of the wild-type and the OxyR strains when plates were exposed to atmospheric oxygen for different periods of time. As can be seen, the OxyR- mutant exhibited a decreased ability to survive under air, compared to the wildtype strain.
t54 Table 5.2 Survival of P. gingivalis W50 wild-type and OxyR- mutant
after exposure to air.
Timeu Wild-typeb OxyR
th 9l¡t%o 92xzVo
2.5 h 83¡zVo 64xzVo
5h 59¡tVo 34+t7o
10h 39xz7o 4xlVo
20tJ. 34tt7o Less than O.05Vo
25h I6¡q%o Less than O.05Vo
u Time in hours of exposure to air'.
b Results represent the percentage of c.f.u. appearing on plates exposed
to air compared to the control (unexposed) and are presented as the
mean + standard deviation of duplicate experiments with duplicate
samples per experiment.
5.3.2.3 HzOz sensitivity in liquid cultures
Table 5.3 shows typical results of experiments to determine the MIC of
H2O2 for the wild-type and the OxyR- mutant. The optical density of each culture was determined after 16 h of batch growth. This experiment was repeated 3 times.
155 Table 5.3. H2O2 sensitivity of liquid cultures of P. gingivalis W50 wild-type
and OxyR'mutant.
IIzOz Anaerobic a ttAerobictt b Concentration Wild Type OxyR- Wild type Oxy R- (mM) 0 1.80 t.7 5 1.78 1.80 0.05 1.85 1.80 1.80 1.80 0.25 1.80 r.75 r.65 t.45 0.5 1.80 1.60 1.ó5 1.30 1 1.85 1.50 1,60 0.15 r.25 1.85 1.50 1.40 0.45 1.5 1.85 1.50 1.30 0.20 r.t5 1.80 1.35 0.95 0.10 2 1.55 0.85 0.15 0.10
-1 1.55 0.15 0.15 0.10 4 0.30 0.15 0.15 0.10 5 0.30 0.15 0.15 0.10 6 0.15 0.10 0.10 0.10 8 0.15 0.10 0.10 0.10 10 0.10 0.10 0.10 0.10 25 0.10 0.10 0.10 0.10
o Results represent the final ODsoon. of cultures, containing different H2O2
concentrations, incubated under an anaerobic atmosphere of Nz:COz:Hz
(90:5:5) for 16 hours.
b Results represent the final ODsoon. of cultures, containing different H2O2
concentrations, incubated under normal atmospheric conditions for 16
hours, with the tube caps slightly opened.
The above results showed that the wild-type was able to tolerate higher concentrations of H2O2 than the OxyR- strain. The MIC for the wild-type, under completely anaerobic conditions, was between 4 and 6 mM while the MIC for the 'When OxyR strain was between 2 and 3 mM. the cultures were incubated
156 "aerobically" (with the tube caps slightly opened) the MIC for both strains was below these levels, although the wild-type still had a higher MIC value than the
OxyR- strain. The H2O2 MIC for the wild-type was between 1.75 mM and 2 mM while the OxyR MIC was between 1 and 1.5 mM.
Moreover, although very close values around the MIC were examined, the
ODsoon. readings showed that when the concentration of H2O2 was approaching the MIC value there was a partial inhibition of growth. To follow this observation, in one set of experiments the OD was read not only after the normal incubation period of 16 hours, but also after 24 hours. This procedure (data not shown) revealed that the cultures showing a partial decrease in OD after 16 hours were able to reach, after 24 houts, the maximum OD. Therefore, to further follow the behaviour of both strains after a H2O2 challenge, a second experiment was carried out in which the growth curves were determined. Results are presented in Figures
5.5,5.6,5.7 and 5.8.
157 Figure 5.5. Effect of the addition of 0.5 mM H2O2 on the growth of P.
gingivølis W50 wild-type (W-T) and OxyR- mutant
1.8 1.6 1.4 1.2 -+ W-T (OmM) ODseon* 1 ---+-- OxyR (OmM) 0.8 HzOz W-T (0.5 mM) added 0.6 +e- OxyB (0.5mM) 0.4 0.2 0 0 10 20 30
Time (hours)
Figure 5.6. Effect of the addition of 0.75 mM H2O2 on the growth of P,
gingivalis W50 wild-type (W-T) and OxyR-mutant
1.8 1.6 1.4 1.2 -+W-T (0mM) ODsoon^ 1 -+OxyR (0mM) 0.8 Hzoz W-T(0.75 mM) added 0.6 ++OxyR-(O.75mM) 0.4 0.2 0 0 5 10 15 20 25 30
Time (hours)
158 Figure 5.7. Effect of the addition of 1 mM HzOz on the growth of P.
gingivalis W50 wild-type (W-T) and OxyR-mutant
1.8 1.6 1.4 1.2 +W-T (OmM) 1 -r-OxyR (0mM) oDsoont 0.8 W-T (1 mM) HzOz 0.6 added ìtOxyR-(1mM) 0.4 o.2 0 0 5 10 15 20 25 30
Time (hours)
Figure 5.8. Effect of the addition of 2 mM HzOz on the growth of P.
gíngivalis W50 witd-type (W-T) and OxyR- mutant
1.8 1.6 1.4 1.2 oDsoont +W-T(OmM) HzOz 1 t*OxyR-(OmM) added 0.8 W-T (2mM) 0.6 ++O> Time (hours 159 From this series of Figures it can be seen that both the wild type and the OxyR- mutant exhibited similar growth curves in the absence of HzOz. As seen in Fig. 5.5, the addition of 0.5 mM HzOz to the growth medium had no effect on the wild-type but retarded the growth of the OxyR strain. Interestingly, about eight hours after the HzOz addition, a recovery of the OxyR- strain was observed and exponential growth was resumed. As shown in Fig. 5.6, the addition of 0.75 mM H2O2 to the growth medium slightly decreased the growth of the wild type for approximately 1-2 hours, after which logarithmic growth continued, reaching stationary phase at almost the same time as the control. In contrast, this concentration of H2O2 stopped the growth of the OxyR- strain for about 15 hours, after which, logarithmic growth resumed. Fig. 5.7 shows that at 1 mM HzOz, the wild-type growth curve began to show a significant change, but the microorganism was still able to recover. In contrast, the growth of OxyR strain stopped completely and did not recover during the time the cultures were monitored. Two mM H2O2 stopped completely the growth of both strains. In summar], these results demonstrated that the OxyR mutant is less able than the wild-type to cope with H2O2 and explain why, in the first experiment, where the MICs were determined after 16 hours of growth in the presence of H2O2, a partial inhibition, at concentrations close to the MIC, was observed. However, it is interesting to observe that around their own MIC, both strains had the ability to recover and resume logarithmic growth. These results suggest then that the OxyR- mutant still possesses certain non-OxyR-dependent mechanisms for the detoxification of }lzO z. r60 5.4 Northern blot analysis of the expression of AhpF-C in P. gingivalís W50 wild-type and OxyR- mutant Among the proteins modulated by OxyR in other microorganisms, is AhpF- C (Rocha et a1.,2000), the alkyl hydroperoxide reductase. The expression of AhpF-C-in P. gingivalis, under anaerobic and oxygenated environments was studied previously (Chapter 4). The ahpF gene was seen to be transcribed as a single RNA molecule with ahpC, increasing slightly under oxygenated growth. In the following experiments, the expression of AhpF-C in P. gingivalis wild-type and OxyR- strains under anaerobic conditions and after treatment with H2O2 was studied. 5.4.1 Methods P. girtgivalis W50 wild-type and OxyR- mutant were grown in batch culture in BHI medium supplemented with 0.5 mg L-r haemin and 5 mg L-l cysteine. At mid-exponential phase each culture was divided in two and transferred to new flasks, one of which was treated with HzOz, while the other was left untreated. An H2O2 concentration of 0.75 mM was used for the wild-type, while 0.5 mM was used for the OxyR- mutant. After 2.5 hours of incubation the cells were used for total RNA isolation, performed as described previously in Section 4.4.1. Probe preparation, northern hybridization and detection were also performed as described previously in the same section. The primers described earlier (Section 4.6.1), were used to amplify a 900 bp fragment of the ahpF that was used for hybridization against P. gingivalis RNA from the wild-type and OxyR', HzOz treated and untreated, cell cultures. Hybridization with the ahpF-specific probe was performed 16t overnight at 42"C, followed by two 5 min washes at room temperature in 2 x SSC buffer containing 0.17o SDS and two 15 min washes at 42"C in 0.5 x SSC containing 17o SDS. 5.4.2 Results Northern analysis of P. gingivalis total RNA from the two strains with a DlG-labeled ahpF-specific probe is shown in Figure 5.9. A transcript of approximately 2.2 kb, containing ahpF and ahpC, appears in the wild-type HzOz- treated and untreated cells, while no transcript was detected in any of the lanes containing RNA from the OxyR- strain. The intensity of the band in the wild-type RNA, however, did not increase following treatment with 0.75 rnM HzOz lor 2.5 hours. 162 Figure 5.9. Northern blot analysis of AhpF mRNA in P. gingiv¿lis wild- type and OxyR- mutant, exposed or not toH2O2. (a) (b) 34 t234 Kb 23S 2.1 - 1.8 16S 1.5 I (a) Ethidium bromide stained membrane to confirm equal loading. (b) Northern blot with AhpF DlG-labeled probe Lane 1. V/ild-type grown anaerobically. Lane 2. Wild-type after treatment with HzOz. Lane 3. OxyR- grown anaerobically. Lane 4. OxyR- after treatment with HzOz. 5.5 NADH oxidase activity in wild-type and OxyR-cell extracts As discussed before in Chapter' 4, AhpF-C is able to act as an NADH oxidase in the presence of oxygen and in the absence of HzOz. Therefore, to evaluate to what extent AhpF-C contributes to the NADH oxidase activity present tn P. gingivalis cell exttacts, assays for NADH oxidase activity were carried out on the wild-type and the OxyR- mutant. As northern blot analysis showed that, unlike 163 the wild-type, the mutant did not constitutively express any AhpF-C, assays for NADH oxidase were performed on cells grown in batch culture under anaerobic conditions. 5.5.1 Methods P. gingivalis V/50 wild-type and OxyR strains were grown in batch culture in BHI medium supplemented with 5 mg L-1 of haemin and 0.5 g Lr of cysteine. Overnight-grown cultures were used as the inoculum for 200 mL broths that were incubated at 37"C, under normal atmospheric conditions with the flask lids tightly closed until mid-exponential phase (OD-soo- 1) was reached. Cell extracts were prepared in duplicate as described previously in Section 4.1.1.1. NADH oxidase activity was assayed following the methods of Higuchi (1992), as previously described. Two assays were done per cell extract and the experiment was repeated twice. 5.5.2 Results Figure 5.10 shows the NADH oxidase activity of the P. gingivalls W50 wild-type and OxyR mutant when grown in BHI medium under anaerobic batch culture conditions. As can be seen, NADH oxidase activity in OxyR- cell extracts was reduced by approximately 25Vo compared to the wild-type. 164 Figure 5.10 NADH oxidase activity of P. gíngivøl¿s W50 wild-type and OxyR'cell extracts. 30 25 NADH 20 oxidase actlvlty 15 (units mg protein-r) 10 5 0 Wild-type OxyR- 5.6 Summary of results and discuss¡on This Chapter describes the initial characterization of the first transcriptional regulator of the oxidative stress response identified in P. gingivalis. No homologues to SoxR-S, PerR, OxyS or OhrR were found in the P. gingivalis genome database. The only transcriptional regulator identified was that belonging to the LysR-type family of regulators, OxyR. After construction of an OxyR mutant, it was confirmed that the ORF encoding for the putative OxyR, might actually have a role in HzOz-detoxification and the resistance of P. gingivalis to 165 exposure to air, as the OxyR mutant exhibited a decreased ability to survive under both situations when compared to the wild-type. Although no complementation was carried out on the OxyR strain, the insertion of the Tetracycline cassette, utilized in the present study, has been shown to produce no further alterations in P. girtgivalis other than conferring resistance to the antibiotic (Dashper et al., 2O0I; Veith et al., 2OO2). Moreover, the growth curves of the OxyR mutant and the wild-type were identical and both strains grew in a similar way on anaerobic blood agar plates, acquiring the pigmented appearance at the same time and showing similar colony morphology. Apart from the OxyR of the aerotolerant B. fragilis (Rocha et a\.,2000), no other OxyR regulons have been characterised in obligatory anaerobic microorganisms. In contrast to P. gingivalis, B. fragilis OxyR mutants are highly sensitive to hydrogen peroxide killing, but their aerotolerance is only marginally decreased. In P. gingivalis, however, OxyR appeared to play a role in both situations, although mutants were still able to survive a certain amount of oxidative stress from both sources, which suggested the existence of other non-OxyR- dependent anti-oxidant systems operating in P. gingivalis. One of the genes that OxyR has been shown to control in other bacteria, is the alkyl hydroperoxide reductase (ahpF-C) (Rocha et a1.,2000). In the present study, P. gingivølls expressed, constitutively (in anaerobic continuous culture), the mRNA molecule containing both ahpF and ahpC, which increased slightly when the cells were grown in the presence of oxygen. When the expression of ahpF-C was assayed in cells grorwn in batch culture, treated and untreated with H2O2, rt was found that these cells also expressed, constitutively (anaerobically), this mRNA molecule. However, no significant differences in expression were found in 166 the cells treated with HzOz. In contrast to the results with the wild-type, the OxyR- mutant did not express ahpF-C either constitutively or after treatment with HzOz. Constitutive expression of AhpF-C has been found to occur in mutated strains of other microorganisms. For example, Rocha and Smith (1998) isolated and characterised peroxide-resistant mutants of B. fragilis that exhibited constitutive expression of AhpC and other proteins involved in the response to HzOz.It was then found that the cause of the deregulation of these proteins was a mutation in the o.r7R regulon at codon 202 (ggt to gat), changing a glycine (G) to an aspartate (D). The authors proposed that this mutation might have caused a conformational change in OxyR, leading to its permanent activation, thereby promoting the constitutive expression of the proteins (Rocha et a\.,2000). Other types of base substitutions have been found to convert E. coli wild.type cells to a permanently activated OxyR phenotype (Kuilik et aL.,1995). Half of the mutations responsible for such a phenotype were found to occur near the critical Cys-199 residue. The sequences of B. fragilis and P. gingivalis OxyR were the most closely related to each other when the phylogenetic relationships of 22 bacterial OxyR proteins were established (Rocha et al., 2000). After aligning the sequence of OxyR (PG 0242) from P. gingivalis V/83 with that of B. fragilis (Fig. 5.1 1), it was found that the same amino acid substitution that occurs in the B. fragilis mutants making them a permanently-activated OxyR phenotype, was present in the P. gingivalis sequence. Noteworthy, is the strong conservation of this particular region, containing as it does the two critical cysteine residues (Cys-199 and Cys- 208). The amino acid at position 202 rs the only difference in this stretch containing 20 otherwise identical amino acids. This amino acid substitution is the 161 product of the same nucleotide bases being changed, as in the B. fragilis mutants (namely a guanine for a thymine). Some OxyR sequences from other microorgansims (e.g. E. coli) carry an aspartate at position 202, which does not translate into permanent activation of the protein. However, it is likely that P. gingivalis OxyR, being so closely related to B. fragilis OxyR, might behave in a similar manner to it. Perhaps the same conformational change that occured in B. fragilis H2O2-resistant mutants occurred pelmanently in P. gingivalis, a more oxygen-sensitive anaerobe, thus ensuring a permanent change in the OxyR conformation to its oxidized state. This could explain then why ahpF-C is constitutively expressed in the wild-type and why the OxyR- mutants do not express it under any condition. This is an interesting finding, as no other transcriptional legulators of the oxidative stress response were found in P. gingivalis. This could, therefore represent an evolutionary modification to compensate for the lack of other defences to oxidative stress. On the contrary, B. fragilis possesses catalase, an enzyme not commonly found in anaerobes, and this could account for most of this microorganism's tolerance to oxidative stress, thus not requiring a permanently activated OxyR. However, whether or not the OxyR from P. gingivalis is constitutively maintained in its oxidized state would require experimental confirmation. Putative proteins, apart from AhpF-C, that might be regulated by P. gingivalis OxyR could include an exodeoxyribonuclease and a single-strand DNA binding protein, the genes for which are found in a "head-to-tail" arrangement in the P. gingivalis chromosome with o-rryR (Fig. 5.I2). It is common for OxyR- regulated genes to be located adjacent to oxyR (Rocha et a1.,2000); but other 168 genes, not located in close proximity, could also be regulated by it, as is the case with ahpF and ahpC. Moreover, as AhpF-C was not constitutively expressed in the OxyR strain, and this protein is capable of NADH oxidase activity in the absence of H2O2 (Niimura et a1.,2000), a comparison of the activities present in the OxyR strain and wild-type was made in order to determine the extent to which AhpF-C was contributing to the NADH oxidase activity measured in P. gingivalis cell extracts. These experiments showed that the mutant had reduced (by 25Vo) NADH oxidase activity, suggesting that this was the contribution of AhpF-C. Therefore, the majority of the activity measured in cell extracts might coruespond to the NADH oxidase activity displayed by the 4-hydroxybutyryl-CoA dehydratase, as indicated by the purification procedures shown in Section 4.2. 169 Figure 5.LL. Sequence alignment of OxyRs from the aerotolerant anaerobe Bacteroides fragilis and the obligatory anaerobe P. gíngivalis. The amino acid difference at codon 202, possibly conferring a permanently oxidised state to P gingivalis OxyR is indicated by the box. P. gingiva-lis MN I QQLEY IAALDKFRHFAKAADYCNVSQPTLS TMI LKLEEELGAKLFDRSRQP I E PTS T B. fragilis MT T QQLEY I LAVDQFRHFAKAAEYCRVTQPTLSAMI QKLEDELGVKLFDRSMT PVCPlAT * ******* *.*.********.** *.*****.** ***.*** *****+ *. **.* P. gingivalis GALVVS QAKQf LYDLNS I TRI I EEEQQSLTGRLNIAVLPl TAPYLLPRVFP IWKKELAGL B. fragilis GKKVLEQARKI LSEVICVKE I I SEEQHSLSGTFRLAVLPT IAPYLLPRFFPQLMEKYPDL * *. **,.** . ** ***.**.* . '************* ** * P. gìngivalis E I HVS EMQTS RCLAS LLSGE I DMAI TASKAETEGLEDDLLYYEE FLGYVSRCE PLFEQDV B. fragilis D T RVMEMKT P D I RKALLTGEADAAI IASMLDDAALTEETLFYEQFLGYVS KKE PL FKHDV .*.* **.*, .**.** * ***** : .* :. *.**.******. ****..** P. gingivalis IRTTEVNPHRLWLLDE cHcrtobr. VRFCQMKGLHERQTAYSGGSMEAFMRLVES GQGI T F B. f ragi-lis I RTS DVTGERLWLLDE GHCFd.GhL VRFCQMEAVKI SQMAYRLGSMET FMHMVES GKGI T F ***,.* ******* *****u** ******, .. * ** ****.**..****.**** P. gingivalis I PQLTVEQLS PSQKELVRP FGMPRPVREVRLAVRQDYS RRKLREQL I GLLRSAVPS DMHK B. fragilis I PELAVMQLS EEQKELVRP FAI PRPTRQI VLVT RKDFI RTS LLQVLKEE I QAAVPKEMLT **.*,* *** *+****** .*** *.. * *.*' * * . * ...*** '* P. gingivalis LQTGQHLA B. fragilis LQAVQCLV 170 Figure. 5.1-2. Genetic arrangement of the region containing oxyR in P. gingivalis. (exoA = exodeoxyribonuclease; sså = single stranded DNA-binding protein). oxyR t7t 6. A comparison of the tolerances to oxygen of P. gingivølis and Fusobacterium nucleøtum and their interactions as a continuous aerated co-culture 112 The studies presented so far have focused on the oxidative stress response of P. girtgivalis, demonstrating that, although it is an obligatory anaerobic microorganism, it possesses a limited ability to cope with oxidative stress. As mentioned earlier in the Introduction, F. nucleatum, another Gram-negative anaerobe, is regarded as a key organism for dental plaque maturation (Kolenbrander et al., 1989). F. nucleatum is found in higher numbers than any other anaerobic species in supra-gingival dental plaque (Ximenez-Fyvie et al., 2000) and in healthy gingival sites (Moore and Moore, 1994) and it has been proposed that it precedes, and is required for, the colonization of P. gingivalis (Socransky et al., 1998). Thus, it seems that F. nucleatum is able to maintain high population numbers in environments likely to possess a high oxygen concentration, supra-gingival plaque for example, where P. gingivalls and other anaerobes are not so prominent. Therefore, the aim of the following experiments was to evaluate the tolerance to oxygen of F. nucleatunt, in comparison to P. gingivalis, and to investigate the possibility that a different oxygen-tolerance is one of the factors that allows F. nucleatunt Lo colonise earlier and in higher numbers than other anaerobes. The interactions of the two microorganisms in co-culture under oxygenated environments were also studied. \73 6.1 Comparison of the tolerances to oxygen of P. gingivalis and F. nucleatum when grown under the same conditions as continuous, aerated monocultures 6.1.1Methods 6.1.1.1 Microorganisms and maintenance of the strains Fusobacterium nucleatumType strain ATCC 10953 and P. gingivalis W50 were used for all experiments and maintained, short-term, on anaerobic blood agar plates incubated at37"C in an atmosphere of Nz/HzlCO2 (90:5:5) 6. 1. 1.2 Growth conditiotts The microorganisms were grown independently as monocultures under continuous culture conditions in BM medium (Shah et al., 1916) supplemented I with 5 mg L of haemin and 0.5 g Ll of cysteine. Growth conditions for both microorganisms were the same as described previously for P. gingivalis in Section 2.2. A dilution rate of D=0.069 h-r lto=16 h) was used in all experiments; temperature was controlled at 36"C and pH maintained at 1.4. The cultures were sparged with the appropriate gas mixture and the En was constantly monitored. Culture purity was checked daily by Gram-staining and cell viability was measured, at steady-state, by viable counts. Under all conditions, steady-state was achieved after about seven generations, as evidenced by sustained stability of the culture E¡ and cell viability. Various culture parameters were then assayed daily for up to 9 days. 114 6. 1. 1.3 Gaseous atnxospheres tested In order to compare the tolerances to oxygen, as monocultures, of the two microorganisms, growth was always initiated under an anaerobic environment (N2/H2lCOz, 90:5:5). At steady-state the incoming gas-mixture was then adjusted to contain increasing oxygen concentrations until wash-out occurred. 6.1.1.4 NADH oxidase and SOD activity of F. nucleatum NADH oxidase and SOD activities of F. nucleatum grown under the different gaseous envilonments were assayed in cell extracts as described previously for P. gingivalis (Section 4.1). SOD activity of F. nucleatum was further analysed by the method of Ilouno et al. (1996). Briefly, the method consisted of measuring the enzyme activity by inhibiting the reduction of nitroblue tetrazolium (NTB). The reaction mixture (3 mL) contained 67 mM potassium phosphate buffer (pH 7.8), 6.6 mM EDTA plus 15 pg ml--t KCN (pH 7.8), 0.05 mM NTB (Sigma), 2 ltll4 riboflavin and the volume equivalent to 1 mg protein extract. The same detection method was utilized to measure the SOD activity of F. nucleatum cell extracts in isoelectric focusing gels (EF), pH 3-10 (Novex). The equivalent of 15 pg of protein from each cell extract was diluted in 2 volumes of Novex sample buffer, loaded on the gel and then run following the manufacturer's instructions. Gels were then placed in a solution containing2.45 mM NTB that was replaced after 20 min with a solution containing 0.5 mg of riboflavin in 0.036 M potassium phosphate buffer (pH 7.8). Gels were incubated in this solution for 15 min and then illuminated with visible light until clear bands appeared on the gel, 175 contrasting against a dark background. Commercial SOD (Sigma) was used as a control. 6. 1. l. 5 Fermentation end-products Fermentation end-products from each stage of growth of F. nucleatum wete assayed using an HPLC system, as described previously in Section2.5. 6.1.2 Results The growth parameters for F. nucleatum and P. gingivalis, when grown as independent continuous monocultures in the same medium but under different oxygenated conditions, are presented in Table 6.1. The levels of NADH oxidase and SOD activities and their inducibility under different gas conditions are compared. The amounts of acetate and butyrate, the dominant metabolic end- products of F. nucleatum and P. girtgivalis, are also presented. It should be noted that some of the data for P. gingivalls were presented previously in Chapter 2. 116 Table 6.L Comparison of the growth parameters, anti-oxidant enzyme activities and metabolic end-products of F. nucleatum (F.n) and P. gingivalis (P.g) when grown under different oxygenated environments. Gas E¡ (mV) " Viable counts OD (soon,n) NADH oxidase" SOD. Acetateo Butyrated Phase Log16 c.f.u. mL-1b P.g F.n P.g F.n P.g F.n P,g F,N P.g F.n P.g F,N P.g F.n Nz/COz -507 -490 9.94 8.60 1.43 t.20 5.5 t90.2 8.9 0.9 7.3 8.5 9.0 13. I (+0 (+0. (+0.2) (+0.9) (+0.3) (+0.4) (95:5) (+0. I 2) (10. r4) (+0. t8) (+0 12) (r0.s) (+3.3) 6) l) Nz/COz/Oz -398 -408 9.05* 8.15 0.98x 1.25 9.8x 600.4+ 10.8x r.4 13.0x 1 1.3x 5.6'4 8.7x (10.2) (+0 (10.3) (75:5:10) (+0. I 8) (+0 I l) (1o.os) (+0. I 0) (rl e) (r4.8) (+0.7) (+0.3) (+0. l) 4) Air/COz wo" -39r wo 8.90* wo 1.50x wo 627.8* wo l.6x v/o 13.5x wo 7 .J'r (95:5) (+0.08) l+o 05\ r+5 gì (r0. r) (+0.3) (r0.ó) Air/COz/Oz wo -274 v/o 8.69 wo t.t 5+ wo 571.2+ wo 1.8x wo 2r.9* wo 4.0x (80:5:15) (+0 t0) (+0.08) r+5 ìl (+0.1) (+0. l) {+(l 3l aE¡, values represent Lhe mean of at least three separate determinations that differed less than 57o. b n=3 for viable counts. " NADH oxidase and SOD activities in units mg protein-r (n=4). One unit of NADH oxidase activity was defined as the amount of extract that catalised the oxidation of I nmol of NADH min-r . One unit of SOD was defined as the amount of extract that produced a 50Vo decrease in the rate of reduction of cytochrome c. d End-p.oducts in mmol (g cells dry weight)-t (n=4). " WO= the culture did not reach steady-state and wash-out occurred. xP < 0.001 for results under each aerated condition compared to anaerobic growth. nl As can be seen in Table 6.I, F. nucleatunt was able to maintain steady-state even when a gaseous atmosphere containing 35Vo oxygen was bubbled through the culture. Above this oxygen concentration wash-out occulred. In contrast, P. gingivalis was only able to maintain steady-state until an oxygen concentration of lOTo was applied, above which it did not survive. The viability of F. nucleatum clult:ures was not diminished by any of the aerated conditions tested. In fact, an increase in cell numbers was observed when oxygen was introduced into the system. In contrast, P. gingivalis, although able to maintain steady-state under some of the gaseous atmospheres tested, decreased its population numbers when oxygen was introduced. An interesting observation was thaT F. nucleatum cultures, when growing under highly oxygenated environments, formed significant biofilms over the chemostat inserts. A tendency to form biofilms was also observed in P. gingivalis cultures during wash-out, and this phenomenon seemed to be the responsible for the slow disappearance of the cells - instead of obeying wash-out kinetics - that occurred when the oxygen in the system exceeded each microorganism's tolerance. Interesting results were also seen when NADH oxidase and SOD activities of the two bacteria were compared. Both F. nucleatum and P. gingivalis "anaerobic" cell extracts contained NADH oxidase activity but, in F. nucleatum, the activity was around 35 times higher than in P. gingivalls. In contrast, SOD activity in P. gingivalis was 10 times higher than in F. nucleatum. Due to the low SOD activity present in F. nucleatum cell extracts, further assays for SOD activity were carried out. Cuvette assays by the method of Ilouno et al. (1996) showed that all samples contained SOD activity (data not shown). Gel assays further confirmed these results and are shown in Figure 6.1. 178 Acetate and butyrate were the fermentation end-products detected in highest concentrations in culture supernatants of both microorganisms. Their concentrations under the different aerated environments are also shown in Table 6.1. In both microorganisms, acetate increased in the presence of oxygen while butyrate decreased as the environments became more oxygenated. In summary, these results showed that the tolerance to oxygen of P. gingivalis is lower than that of F. nucleatunt, which was seen to contain significantly higher levels of NADH oxidase activity and lower levels of SOD activity. Figure 6.1 SOD assay in an IEF get (pH 3-10) for ,F'. nucleatum cell extracts when gro\ryn under different aerated conditions. I ) 3 4 l. Control, commercial SOD. 2. F. nucleatum grown under an anaerobic atmosphere of N2/CO2(95:5). 3. F. nucleatunt grown under an oxygenated atmosphere of N2/CO2IO2 (75:5:10). 4. F. nucleatutn grown under an oxygenated atmosphere of Air/CO2 (95:5). 119 6.2 The survival of P. gingivølis in oxygenated environments when grown as a continuous aerated co-culture with F. nucleatum As the previous experiments demonstrated that F. nucleatuz possessed a higher tolerance to oxygen than P. gingivalis when grown under the same oxygenated environments, it seemed possible that F. nucleatum could support the growth of P. gingivalis tnder oxygenated conditions in which the latter would not survive as a monoculture. To investigate this hypothesis, the microorganisms were grown together as a mixed continuous co-culture under different oxygenated atmospheres. 6.2.1 Methods The ability of F. nucleatum to support the growth of P. gittgivalis tnder aerated conditions was determined by growing both microorganisms as a continuous co-culture under the same gaseous atmospheres described in the previous section for the P. gingivalis and F. nucleatur,n continuous monocultures. The same medium was utilized, except that the concentration of haemin was dropped to 0.5 mg L-l to decrease further the oxygen tolerance of P. gingivalis (this phenomenon was demonstrated in experiments presented in Section 2.3). The same dilution rate (D=0.069), pH Q.Ð and temperature (36'C) were employed, and growth parameters were assayed at steady-state, as described previously. Vortex mixing of co-culture samples, prior to viable counting, was carried out to disperse co-aggregated cells without affecting cell viability. Blood agar plates, with and without 50 pg ml--l of Kanamycin, were used for viable counts of P. 180 gingivalis and F. nucleatum, respectively. In the Kanamycin-containing medium, P. gingivalis was recovered with IO0% efficiency. Cultures grown under various conditions were also examined by scanning electron microscopy (SEM). For planktonic phase analysis, approximately one mL of cell culture was removed from the chemostat and 5 pL were placed on a micro glass cover slip and allowed to dry. When formation of biofilms occurred over the chemostat inserts, the chemostat was disassembled and a small amount of the biofilm was mechanically removed with minimal disruption and placed directly onto cover slips. All samples were then placed in fixative solution containing 4Vo paraformaldehyde, I.25Vo glutaraldehyde and 47o sucrose in PBS, followed by post-fixing in l7o OsOa, and dehydration with increasingly concentrated ethanol solutions. Samples were analysed using a Philips XL30 Field Emission Scanning Electron Microscope. 6.2.2 Results Previous results had shown that when P. gittgivalis was grown in medium containing 0.5 mg L-r of haemin (haemin-limited), under conditions containing I)Vo and 2OVo oxygen, the microorganism was unable to maintain steady state (see results presented in Section 2.3 for P. gingivaiis grown under haemin-limitation). In contrast, the organism survived in co-culture with F. nucleatum in gaseous environments containing these oxygen concentrations (Figure 6.2). Interestingly, as was seen previously in monoculture, a statistically significant increase in the population of F. nttcleatltm was observed in the co- cultures when oxygen (IOVo) was introduced into the system. 181 From Figure 6.2, it can also be seen that the E1, values for the co-culture changed only slightly from anaerobic conditions to llVo Oz. However, the culture E1, increased to about -150 mV under 20Vo oxygen. Daily Gram stains and SEM analysis of the planktonic phase of the culture showed that F. nucleatum cells increased in length as the culture became more oxygen stressed, an observation recorded also for F. nucleatum monocultures. F. nucleatum cells grown under 20Vo ox!{en were at least five times longer than cells grown under anaerobic conditions (Figure 6.3). Moreover, Gram stains and SEM analysis (Fig. 6.4 a, b and c) of the planktonic phase of the culture showed a close physical association between P. gingivalis and F. nucleatum when growth occurred under oxygenated environments, indicating that co-aggregation might be triggered by the stress caused by the presence of oxygen. Although co-aggregating and non-co- aggregating cells were observed in Gram stains of the planktonic phase of the culture under all conditions, the proportion of co-aggregating cells increased at higher oxygen concentrations. In cultures gassed with 20Vo oxygen, large planktonic aggregates were observed and biofilms formed around the growth vessel inserts. SEM analysis of these biofilms showed that the most obvious microrganism on the surface of the biofilms was F. nucleatum, forming an intricate network (Figure 6.4 d). However, viable counts performed on some of these biofilms revealed that P. gingivalis was also present (data not shown). 182 Figure 6.2. Chemostat co-cultures of P. gingivalis and F. nucleatum grown under anaerobic and oxygenated environments. 10 E F. nucleatum I 8P. gingivalis Viable counts Log16 c.f.u. ml-' I (n=5) 7 6 5 Gas Phase -+ Nz/COz (95:5) N2/CO2/O2 Air/COz (95:5) (85:5:10) En=-490 mV Er'=-43 l mV E¡= -158 mV * Viable counts of that microorganism are significantly different to those under anaerobic conditions (P<0.001 ). r83 Figure 6.3. Gram stains of the planktonic phase of a continuous co-culture of P. gingivalis (Pg) and F. nucleatu¡n (Fn) gro\ün under anaerobic conditions (A) and under 207o Oz (B). 1000x magnification. & î tÊ 184 Fig. 6.4 Scanning electron micrographs of planktonic cells (a, b and c) and a biofilm (d) of F. nucleatum and. P. gingivalís grown as a continuous co-culture under 2O7o O,. a b 185 Figure 6.4 (continued) c 2um d 186 6.3 CO2 requirements of P. gingivalis and F. nucleatum During the experiments presented previously, it was noticed incidentally that F. nucleatum did not require the addition of CO2 for growth in the chemostat, while COz seemed to be essential for P. gingivalis. Thus, in the following experiments, this observation was further investigated to explore the possibility that F. nucleatum could supply COz for the growth of P. gingivalis. 6.3.1Methods To evaluate the CO2 requirement of the two microorganisms, each was grown as a continuous monoculture in a Nz/COz (95:5) atmosphere and, when steady state was achieved, COz was excluded from the gas mixtule. The same medium and growth conditions used for all the experiments presented so far in this Chapter were once more utilised. To determine whether P. gingivalis could survive in a CO2-depleted environment, relying only on the amount of COz produced by its own metabolism, a closed environment was used, rather than a chemostat in which the gaseous atmosphere is continually replaced. For these experiments, P. gingivalis was grown in batch culture; two anaerobic jars were used for incubation at 37"C, one gassed with Nz/COz (95:5) and the other with Nz alone. P. gingivalls cells growing at steady-state in continuous culture under an atmosphere of Nz/COz (95:5) were used as the initial inoculum. BM medium was used for all experiments and media were inoculated to produce an initial OD5oonn' of 0.45, in a total volume of 50 mL. At late exponential phase the ODseo was recorded, the pH was measured and the appropriate amount of culture, to produce an initial ODsoo of 0.45, was transferred 187 to fresh medium. These broths were then incubated under the appropriate gaseous conditions. The ability of F. nucleatum to support the growth of P. gingivalis in the absence of COz, under anaerobic and aerated conditions was determined by growing them in continuous co-culture under the same gaseous atmospheres as described above (Section 6.2),but COz was excluded from the gas mixture. 6.3.2 Results and discussion The carbon dioxide requirements of anaerobic bacteria have been studied for decades (e.g. Reilly, 1980) as COz is required for the bio-synthesis of one- carbon units that occupy specific positions in purines, methionine, arginine and other important building blocks in the cell. As seen in Figure 6.5, the exclusion of COz from the gas mixture resulted in wash-out of a monoculture of P. gingivalis but not of F. nucleatum. This certainly occurred because, in a continuous culture system, COz cannot accumulate to a threshold concentration unless it is continuously replenished. This effect might be flow-rate dependent and perhaps at a higher flow-rate F. nucleatum would not have survived as the CO2 produced endogenously by the microorganism "washed out". The need of P. gingivalis for exogenously added CO2, was also seen in batch culture experiments, under a closed anaerobic atmosphere. As shown in Table 6.2, the microorganism did not even survive a single transfer in a COz- depleted atmosphere, although alarge inoculum was utilised for the transfer. Figure 6.6 shows that, in continuous culture, F. nucleatLtm was able to satisfy the requirements of P. gingivalis for COz, since the exclusion of COz from the gas mixture did not have a marked effect on P. gingivalis population sizes 188 when grown in co-culture with F. nucleatum. This contrasts with the failure of P. gittgivalis monocultures to survive without CO2 (Fig. 6.5). This ability of F. nucleatunt to support the growth of P. gingivalis in the absence of CO2 was also maintained under oxygenated atmospheres. Thus, F. nucleatum appears able to produce sufficient COz for it to maintain the growth of P. gingivalis tn the continuous culture system. Figure 6.5. Effect of COz on the chemostat growth of F. nucleatum and P. gin giv alis monocultures* CO2 turned off 1.4 + 1.2 1 Optical 0.8 Density (560nm) 0.6 0.4 + P. gingivalis o.2 -+- F. nucleatum 0 5 7 9 11 13 15 17 19 Generations * Th" gas phase until the 72th generati.on was Nz/COz (95:5). t89 Table 6.2. Effect of COz on the batch culture growth of P. gingivalís. Atmosphere oDu OD TRO Nz:COz(95:5) 1.15"10.05 (pH=7.5) r.25 +0.04 N2(100) 1.05+0.05 (pH=1.4) 0.38 +0.06 u ODlsoo nm) wÍtS measured at late exponential phase. Initial OD for inoculum and transfers was 0.45. o OD TR represents the optical density (5ó0 nm) of transferred cultures at late exponential phase, as explained in the Text. " Mean + SD of 3 separate experiments. 190 Figure 6.6 Chemostat co-cultures of P. gingivalis and F. nucleatum grownin the absence of CO2 under increasingly oxygenated conditions 10 E F. nucleatum 9 @ P. gingivalis I 7 6 5 N2 (100) Nz/Oz (90:10) Air (100) Gas phase * V¡able counts of that microorganism are significantly different to those under anaerobic conditrons (P<0.001 ). 6.4 Summary of results and discussion The present studies clearly show that, although F. nucleatum and P. gingivalis are anaerobes, they can each tolerate oxygen, albeit to a different extent. As shown in Table 6.I, F. nucleatum is able to survive in chemostat cultules sparged with proportions of oxygen even higher than the oxygen content of air, 191 while P. gingivalls is not able to survive, in the same system, when the gas phase of the culture contains more than tÙTo oxygen. Leke et al. (1999) demonstrated that F. nucleatum and P. gingivalis would not grow de novo under aerobic conditions. This was almost certainly due to the fact that the E¡ of the uninoculated medium was too high. In the present study it was also found that de novo chemostat growth of both microorganisms could not be initiated under aerobic conditions. However, as shown in Table 6.I, a dense anaerobic culture could grow in the presence of different concentrations of oxygen. If the levels of NADH oxidase and SOD activities in P. gingivalis and F. nucleatum are compared, interesting differences are found. Under anaerobic conditions, F. nucleatum produced 190 units of NADH oxidase activity (mg protein)-r, while the P. gingivalis activity was only 8.15 units (mg protein)-l; a difference of more than 2O-fold. Moreover, NADH oxidase activity in F. nucleatum was stimulated by the presence of oxygen, reaching levels of activity around 600 units (mg protein)-I, while the activity in P. gingivalis, under the highest oxygenated condition, was not higher than 10 units (mg protein;-r. Earlier results in this study showed that the NADH oxidase activity measured in P. gingivalis cell extracts was due to the NADH oxidase-like activity of different enzymes, 4-hydroxybutyryl-CoA dehydratase and alkylhydroperoxide reductase, while no evidence of the expression of a "true" NADH oxidase was found. A similar situation could occur in F. nucleatum, in which different enzymes might contribute to the NADH oxidase activity measured. However, the higher activity detected in F. nucleatum might suggest the possibility that, in this microorganism, the NADH oxidase is functional. In fact, two ORFs encoding for putative NADH oxidases were found in the preliminary genome sequence of F. nucleatum ATCC 192 10953 (http://www.hgsc.bcm.tmc.edu). Moreover, a gene encoding for a putative alkyl hydroperoxide reductase was also found. Kapatral et al., (2002) also reported in their publication of the genome sequence of F. nucleatunt súain ATCC 25586, that this microorganism might use its NADH oxidase for growth in the presence of oxygen. In contrast to the NADH oxidase activity, SOD activity in F. nucleatum was considerably lower than in P. girtgivalls. Thus, in these two microorganisms, SOD activity does not correlate with tolerance to oxygen, which is at variance with previous suggestions for other species (Tally et aI., I97l;Park, et al., 1992).The role of SOD is the detoxification of O2'- with the formation of H2O2 and Oz (McCord and Fridovich, 1969). Therefore, high levels of SOD, without correspondingly high levels of NADH oxidase (to metabolise the O2), could be toxic for anaerobic microorganisms because of the potential generation of large amounts of ROS, a situation that could be occuring in P. gingivalis. When the levels of the dominant end-products produced by the two microorganisms were compared, a similar situation was observed in the presence of oxygen. That is, both microorganisms tended to decrease butyrate production in the presence of oxygen, while the amount of acetate increased with the concentration of oxygen. This might be an indication fhat F. nucleatum utilizes similar pathways to those of P. gittgivalis for the generation of these end-products. Thus, it is possible that the metabolism leading to butyrate formation might be decreased in F. nucleatum becatse of the oxygen sensitivity of 4-hydroxybutyryl- CoA dehydratase. It was also interesting to observe that the viability of F. nucleatum was increased under moderately oxygenated environments. These results could be linked with the greater ability of F. nucleatum to regenerate 193 NADH and divert its metabolism to acetate production; a more efficient pathway in some microorganisms (Stanton and Jensen, 1993). In the recent publication of the F. nucleatum ATCC 25586 genome sequence (Kapatral et a1.,2002), it was noted that the organism's NADH oxidase might be important in accommodating oxidative stress due to the absence of superoxide dismutases or catalases. However, the results of the present study showed that F. nucleatum cell extracts possess low SOD activity. This activity was measured in cuvette assays by two methods, with positive results, and was further confirmed by an assay on an IEF gel. Extracts of F. nucleatum, under all the conditions tested, exhibited SOD activity and different bands with activity were detected in the IEF gels - a situation that is common for this enzyme, which possesses different isoforms (Yesilkaya et aI.,2000). The possibility was also considered that the enzyme assayed, rather than being superoxide dismutase, was a superoxide reductase. The latter is a novel mononuclear iron-containing enzyme, which catalyses the reduction, rather than the dismutation, of superoxide to hydrogen peroxide (Adams et al., 2002).It has been found in the genome of some anaerobic bacteria, which do not possess SOD, and has been proposed as a novel pathway for the detoxification of superoxide without the generation of oxygen. The assays utilized for SOD will not distinguish between this enzyme and superoxide reductase (Jenney et al., 1999). However, after a Blast search, no superoxide reductase homologues were found in the preliminary F. nucleatum ATCC 10953 genome sequence, excluding the possibility that instead of a SOD, the enzyme assayed in F. nucleatum was a superoxide reductase. 194 It is not clear then which enzyme is responsible for the SOD-like activity measured in F. nucleatum. The only study found in the literature in which SOD has been assayed in F. nucleatum cell extracts reported the testing of 3 different F. nucleatum strains. Two of them exhibited no SOD activity, while a Type strain, ATCC 23126, possessed very low levels of activity (Amano et al., 1986). Strain ATCC 10953 (used in the present study) was not included in the study by Amano et al. (1986). Future studies are needed to clarify which enzymes are responsible for SOD-like activity in F. nucleatum. The present results clearly show that F. nucleaturn is able to support the growth of P. gingivalis tnder aerated conditions in which the latter cannot survive as a monoculture. Indeed, the fact that the populations of F. nucleatum increase under oxygen stress indicates its capacity to survive in natural environments that might be partially oxygenated; in contrast to the low tolerance to oxygen of P. gingivalis. Thus, these results suggest that the capacity of F. nucleatum to protect P. gingivalls from oxidative damage might be one of the reasons why there seems to be a close in ylvo association between these two microorganisms (Socransky, et a1.,1998). Additionally, this study shows that the growth of F. nucleatum does no| require a capnophilic environment, while COz seems to be essential for P. girrgivalis. The fact Íhat P. gingivalis survived in co-culture with F. nucleatum without the addition of COz, indicates that the latter is able to satisfy the COz requirement of the former. Possibly, this constitutes another metabolic interaction that explains the close association between the two microorganisms. Other possible interactions include the presence of proteolytic enzymes in P. gingivalis that cleave 19s proteins into peptides and, therefore, increase the energy sources for F. nucleatum, which does not possess high endopeptidase activity (Grenier, 1994). Evidence exists that a biofilm configuration might protect anaerobic microorganisms from oxygen toxicity (Fukuda et a1.,2001). In fact, it has been demonstrated that the respiratory activities of a microorganism are markedly reduced when growth occurs as a biofilm compared to those of cells growing in suspension (Nguyen et aL.,2002). The formation of aggregates, as observed in the present study, might be a strategy used by F. nucleatum and P. gingivalis to overcome high oxidative stress, possibly because of formation of more reduced microenvironments. It was also interesting to observe that when P. gingivalis cultures reached their maximum tolerance to oxygen and started to disappear from the chemostat, biofilms formed over the chemostat inserts. The same occurred in F. nucleatum monocultures and in the co-culture, however, the biofilms formed by the co-culture were more obvious and occurred not only during wash-out, but also when the cultures were growing at high oxygen concentrations. Therefore, it seems that for both microorganisms, the formation of biofilms is triggered by the presence of oxygen. The intricate networks formed by F. nucleatum in the co- culture biofilms are not sutprising, considering the long rod shape of the microorganism, which increases its length in an oxygenated environment. These "nets" might provide P. gingivalis with adequately reduced conditions in which to survive aeration. In conclusion, additional to the known direct potential involvement of F. nucleatum in the disease process (Han et al., 20OO; Yoshimura et al., 1997), this study suggests that F. nucleatum could have an imporlant indirect role in the aetiology of periodontal diseases by supporting the growth of P. gingivalis, and 196 possibly other oral anaerobes, in oxygenated and carbon dioxide-depleted environments. Protection against the deleterious effects of oxygen might be important in both the early stages of plaque development and in periodontal pockets, where microorganisms have "to face" the constant presence of residual oxygen levels. From an ecological point of view, the identification of microorganisms that, like F. nucleatum, support the growth of other periodontopathogens is important, because control of such species might radically alter the pathogenic ecosystem. t9'7 7. General discussion 198 For the anaerobic periodontopathogen P. gingivalis to survive in the oral environment, where it is exposed to a diverse range of oxidative stresses, it requires anti-oxidant defense systems. In the present study, these mechanisms have been investigated and the results obtained assist in our understanding of how this and other obligatory anaerobic microolganisms adapt to the unfavorably oxygenated conditions of the mouth at the different stages of colonisation and disease progress. Understanding such mechanisms might increase our knowiedge of the ecology of dental plaque. In the present Chapter, the mechanisms by which oxygen is toxic to P. gingivalis will be discussed, as well as the anti-oxidant defense systems operating in this microorganism. The interactions of P, gingivalis and F. nucleatum under oxygenated environments will also be reviewed. 7.1 Oxygen toxicity to P. gingivalis Superoxide and hydrogen peroxide are formed inside aerobic cells because of the reaction of oxygen with a solvent-exposed flavin of enzymes such as NADH dehydrogenase I[, succinate dehydrogenase, fumarate reductase and sulphite reductase (Imlay, 1995; Messner and Imlay, 1999). All of these enzymes are members of electron-transport chains and utilise FAD for univalent electron transport reactions the kind of reactions required for the generation of superoxide. Interestingly, it has been shown that iron-sulphur clusters and haems, facile univalent reductants also present in electron transport chains, seem not to react substantially with oxygen (Messner and Imlay, 1999). 199 In anaerobes, it is thought that enzymes also containing FAD are a primary source of ROS, although insufficient evidence exists regarding the biological basis of anaerobiosis. The primary product of the auto-oxidation of flavoproteins is HzOz, but some superoxide is also generated (Messner and Imlay, 1999). The reduced flavins transfer a single electron to oxygen, resulting in the formation of a flavosemiquinone and a superoxide anion. Some flavoenzymes can then undergo a spin inversion before superoxide escapes the active site, allowing the recombination of superoxide and flavo semiquinone radicals and the formation of a peroxy adduct. The HzOz formed is then released by heterolytic cleavage. However, depending on the ability of the enzyme to undergo spin inversion, some superoxide can escape the active site before the adduct can be formed (Imlay, 2002). Thus, as the generation of superoxide radicals and H2O2 is a consequence of the reaction of oxygen with flavoproteins, the amount of these two pro-oxidant species generated inside a bacterial cell might depend upon the enzyme profile of the bacterium and is an important factor in establishing the sensitivity of a microorganism to oxygen (Imlay, 2002). A search of the P. girtgivalis genome sequence for enzymes with the potential to generate high levels of ROS reveals that this organism possesses a putative fumarate reductase; an enzyme involved in the reduction of fumarate to succinate under anaerobic conditions (Iverson et aL.,1999). Fumarate reductase has been shown to be a primary source of ROS in E. coli under aerobic conditions (Imlay, 1995). Furthermore, the present study, in an attempt to purify an NADH oxidase, revealed that 4-hydroxybutyryl-CoA dehydratase is abie to react with NADH, 200 reducing the flavin and generating substantial amounts of HzOz.4-hydroxybutyryl- CoA dehydratase accounts for 15 to 25Vo of the total soluble protein of Clostridium aminobutyricum (Scherf and Buckel, 1993).It is possible that this enzyme is also present in substantial amounts in P. gingivalis.Indeed, butyrate, the end-product of the metabolic pathway in which the dehydratase is involved, is one of the dominant end-products of P. gingivalis metabolism. It is interesting that the dehydratase activity towards its substrate, 4-hydroxybutyryl-CoA, is lost in the presence of oxygen. Thus, the FAD required for this activity would then be "free" to react with NADH, producing FADH2 that would subsequently react with oxygen to generate Hzoz. Exogenously-generated HzOz could also be a source of toxicity to P. gingivalis. Hydrogen peroxide, which diffuses readily into the cells, can be generated in culture media when growth substrates react with oxygen (Imlay, 2002). Exogenous sources of HzOz in the oral environment would include also that produced by some microrganisms in dental plaque, e.g. streptococci of the viridans group (Barnard and Stinson, 1999). The present study also showed that the profile of fermentation end-products of P. girtgivalis altered when the environment was changed from anaerobic to oxygenated conditions. The fact that oxygen is toxic to enzymes involved in the production of butyrate is reflected in the decrease in this end-product under aeration. The cells, therefore, might divert their metabolism to different fermentation pathways, as indicated by the increase in acetate production. In other microorgansims, this shift in end-products in the presence of oxygen has been associated with an increase in NADH oxidase activity, allowing the cells to divert their metabolism to the production of acetate, a more efficient pathway. As NADH 201 oxidase regenerates NAD, the need to utilise the pathway that produces butyrate is avoided (Stanton and Jensen, 1993). However, in P. gingivalis there does not seem to be such a major increase in the consumption of NADH as in other microorgansims (e.g. F. nucleatum). Rather, the decrease in butyrate noted could be attributed to the inactivation by oxygen of enzymes, such a 4-hydroxybutyryl- CoA dehydratase, involved in the production of butyrate from glutamate. The residual NADH oxidase-like activity of the dehydratase (plus the NADH oxidase activity of AhpF-C) might help the cells to increase acetate production, but this would be limited by the available pools of NAD. Therefore, tn P. gingivalis, survival under oxygenated conditions does not seem to be such an efficient process in terms of metabolic yield, as it is in F. nucleatum. Perhaps this is the explanation for the reduced cell yield observed in the former, while the latter increases its cell numbers in the presence of low oxygen levels. 7 .2 Tlne anti-oxidant defences of P. gingivalis In the present study, it has been shown that P. girtgivalis can tolerate and grow in the presence of oxygen in the environment (below normoxic levels), when grown under continuous culture conditions. Moreover, when individual cells are exposed to air, llVo of the population remains viable after25 hours. These results might be an indication of the microorganism's ability to tolerate the constant or transient presence of oxygen in different oral niches. The mechanisms contributing to the oxygen tolerance of P. gingivalis, discussed previously in the Text, are summarised in Table 7.1. Included in the Table is a possible role for ferritin. 202 Figure 7.1 Anti-oxidant protective mechanisms of P. gíngívalís Mechanism Mode of Action Reference p-oxo dimeric haem - Haem (Fe(II)PPD(), after being released from host proteins, sequesters Smalley et al., (1998; layer Oz and binds to the cell in the Fr-oxo form [Fe(III)PPIX]2O. 2000) - The layer serves as a physical barrier against 02. Present study - The layer possesses peroxidase-like activity. Morphological change Unknown, perhaps by this mechanism the cells decrease the ratio of area Present study from a coccoid to a rod exposed to oxygen/cell volume, or alternatively, increase the possibilities of shape biofilm formation. SOD Superoxide detoxification Lynch and Kuramitsu, (teee) Alkyl hydroperoxide H2O2-detoxification Present study reductase (AhpF and AhpC) Rubrerythrin Unclear, protective either through HzO:-detoxification or by sequestering iron Sztukowska et al., (2002) Thioredoxin-linked thiol HzOz-detoxification . The genes for these proteins are present in P. gingivalis peroxidase genome sequence. Experimental confirmation of their role is necessary. RecA DNA repair (indirect evidence) Liu and Fletcher, (2001) OxyR Control of the expression of AhpF-C Present study Permanently activated? Possible control of the expression of other proteins such as Dps, ExoA, Ssb, Fur and Trx Ferritin A putative role decreasing levels of intracellular free iron has been proposed in other microorganisms. Experimental confirmation is necessary. 203 Results in the present study show that P. gingivalis maintains most of its proteolytic capacity in the presence of oxygen. In the periodontal sulcus, P. gingivalis proteinases will aid the release of haemoglobin from erythrocytes and its subsequent degradation to liberate the haem molecule. Some of this monomeric haem, in the reduced state, will be bound immediately to the cell surface in the form of dimers, a mechanism that consumes oxygen (Smalley et al., 1998). The haem that remains in solution, will react with any oxygen in the environment to generate dimers that will also aggregate on the cell surface, providing the cells with a physical barrier to oxidative agents, as presently demonstrated, and also serving as a "catalase-1ike" buffer system fas suggested by Smalley et al., (2000)]. However, when present in supra-gingival plaque, the main role of the proteinases would be nutritional. In this respect, the fact that increased cell- associated Arg-gingipain was found in oxygen stressed cells could be beneficial to the cells in order to release peptides from proteins in the immediate cell environment. No mRNA molecule or protein corresponding to an NADH oxidase (Nox) was detected in P. gingivalis in the present study. An attempt to isolate the NADH oxidase activity present in cell extracts resulted in the purification of a dehydratase that seems to react with NADH, producingH2O2. Therefore, this protein instead of constituting an anti-oxidant system might rather have a toxic effect. In contrast, the expression of mRNA for alkyl hydroperoxide reductase (AhpF and AhpC) was detected under anaerobic and oxygenated conditions and its expression was seen to be dependent on OxyR, as OxyR- mutants did not show any mRNA for the enzyme, either constitutively or under oxygen stress. This finding 204 suggested, then, the possibility that OxyR is permanently activated in P. girtgivalis. Moreover, the same base mutation found in B. fragilis mutants that confers a permanently oxidised state to their OxyR, was found in the P. gingivalis W83 OxyR sequence. Such a permanent state of activation of OxyR might be an evolutionary adaptation. Facultative and aerobic bacteria appeil to have calibrated their OxyR and SoxRS control systems so that they are not activated by the doses of oxidants that are generated during normal aerobic metabolism (lmlay, 1995). P. gingivalis, however, being an obligatory anaerobic microorganism that inhabits an oxygenated environment, might need to ensure that the transcription of OxyR- dependent proteins is permanently activated. The permanently-activated state of OxyR in P. gingivalls does, however, require experimental confirmation. From an evolutionary perspective, it is important to understand that bacteria have evolved to produce the necessary mechanisms allowing them to survive in their natural environment. P. gingivalis might be more sensitive to oxygen than many other microorganisms, including other oral anaerobes such as F. nucleatum. However, its anti-oxidant systems, and perhaps its interactions with other species, appear to be sufficient for persistence in the oral environment until a niche with adequate conditions for growth can be found. Moteovet, once the microorganism has established in the periodontal sulcus, these anti-oxidant systems might still be important, as other sources of oxidative stress, such as polymorphonuclear leukocytes, might be encountered. Furthermore, if P. gingivalis enters the blood stream, oxidative stress defences might be important virulence factors, as the cells will again be challenged by ROS produced by polymorphonuclear leukocytes and other cells of the host immune system; this in addition to the fact that a partial pressure of oxygen from 70 to 95 mm Hg (approximately 7 to I3Vo 02) exists in 205 arterial blood (Park et al., 1992).Indeed, it has been demonstrated that a shift in pH to that found in blood is a stimulus for the induction of protection against oxidative stress in S. gordonll (Vriesema et al., 2000). Therefore, an adequate response to oxidative stress might be one of the factors necessary for interactions of P. gingivalis with the host. Future studies should include the evaluation of the role of AhpF-C, OxyR and other proteins involved in the response to oxidative stress on the in vivo virulence of P. gingivalis. 7.3 F. nucleatum protects P. gingivalis from oxidative damage Of the 300 to 400 species isolated from the oral cavity, only a small group of microorganisms, including F. nucleatum and P. gingivalis, is consistently associated with periodontitis (Socransky and Haffajee, 1994). These two species, and other periodontopathogens, might colonise the supra- and sub-gingival plaque of periodontally-healthy individuals for considerable periods of time prior to disease initiation. Indeed, it has been demonstrated that all suspected periodontal pathogens, most of them obligate anaerobes, occur at a subset of supra- and sub- gingival sites and in healthy subjects; albeit in low numbers and in low proportions (Ximenez-Fyvie et al., 2000). Differences exist among obligate anaerobes in their prevalence and proportions in supra- and sub-gingival plaque. For example, important proportions of F. nucleatum subspecies exist in the supra-gingival plaque of healthy subjects, while P. girtgivalis is found at almost undetectable levels (Ximenez-Fyvie et al., 2000). The factors that might determine the levels of a microorganism in a particular environment might be nutritional, pH- or oxygen-related. Thus, 206 differences in tolerance to oxygen between species might be an important determinant in the process of colonization of dental plaque and in the shifts occurring in the transition from health to disease. Socransky et. al. (1998) demonstrated that bacteria in sub-gingival plaque exist as microbial complexes. Although P. gingivalis and F. nucleatum were identified as forming part of different complexes, a close association was shown to exist between the complexes to which each microorganism belonged. In fact, the members of the P. gingivalis complex were rarely found in the absence of members of the F. nucleaturø complex. The data from the present study provide an explanation for this apparent dependence of P. gingivalis on F'. nucleatum, suggesting that the latter might create the necessary reduced conditions and supply COz for the survival of the former. Cooperation of a similar kind has been suggested to exist between facultative anaerobic species, which were shown to be essential for the survival of anaelobic microorganisms in a lO-membered chemostat community (Bradshaw et a1.,1991). Differences in the tolerance to oxygen between F. nucleatum and P. gingivalis appeared also to exist in a study canied out by Bradshaw et aI., (1991), who tested the ability of a 4-species mixed community of anaerobes to survive in an aerated environment in a two-stage continuous culture system. The four species comprising the community were F. nucleatum, P. gingivalis, Prevotella nigrescens and Veillonella dispar. When the culture was transferred from the first-stage anaerobic chemostat to the second aerated chemostat the death rates were more dramatic than in the present study. That is, in the aerated stage, the viable counts of F. nucleatum wele decreased more than three-fold, while P. gingivalis was not detected. Despite this difference, F. nucleatum was still able to maintain steady- 207 state, in very low numbers, in this community under an atmosphere in which arr was the sole gas phase. P. gingivalis, however, completely disappeared from the culture. Perhaps the differences with the present study, in which F. nucleatum populations were not so markedly affected by the presence of oxygen, could be attributed to differences in growth medium. In the studybyBradshaw et. al.,BM diluted five-fold and supplemented with porcine gastric mucin was used, while in the present study, undiluted BM supplemented with cysteine, a weak reducing agent, was employed. Moreover, a dense culture that was subsequently aerated was currently used, while Bradshaw et. al. utilised a culture that was slowly inoculated into a second aerated vessel. As the capacity of a culture to metabolise oxygen will depend upon cell numbers, a dense culture has more chance of survival in an aerated environment than does a small inoculum with a low number of cells. Thus, it is possible that F. nucleatum,'achieving steady-state at such low numbers, was not able to adequately reduce the environment for the survival of P. gingivalis. Neveltheless, the study by Bradshaw and co-workers indicated that a difference existed in the tolerance to oxygen of F. nucleatum and P. gingivalis. These differences were also shown in the present study when mono-cultures of both microorganisms were compared fol their tolerance to oxygen. A greater ability to deal with oxidative stress could explain the higher proportions of F. nucleatum in supra-gingival plaque compared with the rest of the strict anaerobes. A detailed study of the anti-oxidant mechanisms of F. nucleatum would be an interesting topic for future studies, perhaps allowing us to understand the differences in tolerance to oxygen among anaerobes. It has been demonstrated that biofilms can provide bacteria with a "haven" from the effects of antimicrobial agents, other hostile environmental factors and 208 events such as phagocytosis (Costerton et a|.,1987). The formation of biofilms, as observed in this study, might be a strategy used by F. nucleatum to overcome high oxidative stress, possibly because of the reduced microenvironments inside the biofilm, which could in turn benefit P. gingivalls. Moreover, the change in the cell shape of the two microorganisms (to elongated forms) in the presence of oxygen could increase the possibility of cell contact and promote biofilm formation. Future studies should be aimed at looking for factors that could promote biofilm formation under oxygenated conditions; particularly in the case of F. nucleatum, a key-microorganism in the physical structure of dental plaque due mainly to its co- aggregative ability (Bradshaw et aL.,1998). ln summary, during dental plaque maturation and the associated shifts occurring from health to disease, the interactions between F. nucleatum and P. gingivalis could be described as follows. A higher tolerance to oxygen might allow F. nucleaturø to survive in higher numbers than P. gingivalis in supra- and sub- 'When gingival plaque of healthy individuals. environmental conditions change, e.g. due to undisrupted plaque accumulation, F. nucleatum might rapidly proliferate in moderately reduced environments increasing its numbers and participating in the early stages of disease progress. Indeed, the predominance of F. nucleatum in the sub-gingival plaque of sites showing clinical signs of gingivitis has been reported (Slots and Chen, 1999). Higher numbers of F. nucleatum would consequently lower the redox-potential, promoting the growth of P. girtgivalis. Increased proportions of P. gingivalis might contribute as well to lower the E¡ of the environment, perpetuating the reduced conditions and the progress of disease. In conclusion, the present studies show that, as a single species, P. girtgivalis is able to tolerate certain levels of oxidative stress via mechanisms 209 involving the accumulation of haem on the cell surface, metabolic and morphological adaptations and the expression of a protein (AhpF-C) involved in hydrogen peroxide detoxification. The constitutive expression of this protein appears to be under the control of the transcriptional regulator OxyR, which is important for the survival of the microorganism under oxygenated and hydrogen peroxide-containing environments. In a mixed community, the survival of P. gingivalis, in an oxygenated environment, might be enhanced by the presence of other species more tolerant to oxygen. 210 Bibtiography Abdollahi, H. and Wimpenny, J. W. T. (1990). Effects of oxygen on the growth of Desuffovibrio desulfuricans. J Gen Microbiol. 136: 1025-1030. Adams, M. W. W., Jenney, F. E., Clay, M. D. and Johnson, M. K. (2002). Superoxide reductase: fact or fiction? J Biol Inorg Chem' 7(6):641-652. Altuvia, S., Weinstein-Fischer, D., Zhang,A., Postow, L. and Storz, G. (1997)' Escherichia coli regulatory RNA OxyS gene, complete sequence. http ://www.ncbi.nlm.nih. gov. Amano, 4., Sharmâ, 4., Sojar, H., Kuramitsu, H. and Genco, R. J. (1994). Effects of temperature stress on expression of fimbriae and superoxide dismutase by P orphyromonas gin giv alis. Infect Immun' 62: 4682-4685. Amano, 4., Shizukuishi, S., Tamagawa, H., Iwakura, K., Tsunasavva' S. and Tsunemitsu, A. (1990). Characterization of the superoxide dismutases purified from either anaerobically maintained or aerated Bacteroides gingivalis. J Bacteriol. I72: 1451-1463. Amano, 4., Tamagawa, H., Shizukuishi, S. and Tsunemitsu, A. (1986). Superoxide dismutase, catalase and peroxidases in oral anaerobic bacteria. J Osaka Univ Dent Sch. 26: I8l-I92. Amano,4., Tamagawa, H., Takagaki, M., Murakami, Y., Shizukuishi, S. and Tsunemitsu, A. (1988). Relationship between enzyme activities involved in oxygen metabolism and oxygen tolerance in black-pigmented Bacteroides. J Dent Res. 67(9): 1196-9. Barnard, J. P. and Stinson, M. \ry. (1999). Influence of environmental conditions on hydrogen peroxide formation by Streptococcus gordonii.Infect Immun, 67 (t2): 6558-6564. 211 Bartold, P.M., Gully, N. J., Zilm, P. S. and Rogers, A. H. (1991). Identification of components in Fusobacterium nucleatum chemostat-culture supernatants that are potent inhibitors of human gingival fibroblast proliferation. J Periodont Res. 26: 314-322. Belo, I., Pinheiro, R. and Mota, M. (2000). Response of the thermophile Thermus sp RQ-l to hyperbaric air in batch and fed-batch cultivation. Appl Microbiol Biotechnol. 53(5): 571 -524. Bhogal, P. S., Slakeski, N. and Reynolds, E. C. (1997). A cell-associated protein complex of Porphyromotxas gingivalis W50 composed of Arg- and Lys- specific cysteine proteinases and adhesins. Microbiol. 143: 2485-2495. Blattner, F. R., Plunkett, G., Bloch, C.4., Perna, N. T., Burland, V., Riley, M., Colladovid€s, J., Glasner, J. D., Rode, C. K., Mayhew, G. F., Gregor, J., Davis, N. W., Kirkpatrick, H. .A'., Goeden, M.4., Rose, D. J.' Mau, B. and Shao, Y. (1991). The complete genome sequence of Escherichia coliK-12. Science. 211(5331): 1453 ff. Bolotin,4., Wincker, P., Mauger, S., Jaillon, O., Malarme, K., Weissenbach, J., Ehrlich, S. D. and Sorokin, A. (2001). The complete genome sequence of the lactic acid bacteritm Lactococcus lactis ssp lactis n-I403. Genome Res. 1l(5):131-153. Booth, I. R. (2002). Stress and the single cell: Intrapopulation diversity is a mechanism to ensure survival upon exposure to stress. Internat J Food Microbiol. 78(l-2 Special Issue SI): 19-30. Bowden, G. H. W. and Hamilton, I. R. (1998). Survival of oral bacteria. Crit Rev Oral Biol Med. 9(1): 54-85. 212 Bradshaw, D.J., Marsh, P.D., Allison, C. and Schilling, K. M. (1996). Effect of oxygen, inoculum composition and flow rate on development of mixed- culture oral biofilms. Microbiol. I42(Part 3): 623-629. Bradshaw, D. J., Marsh, P. D., Watson, G. K. and Allison, C. (1997). Oral anaerobes cannot survive oxygen stress without interacting with facultative/aerobic species as a microbial community. Lett Appl Microbiol. 25(6):38s-387. Bradshaw, D. J., Marsh, P. D., Watson, G. K. and Allison, C. (1998). Role of Fusobacterium nucleatum and coaggregation in anaerobe survival in planktonic and biofilm oral microbial communities during aeration. Infect Immun. 66( 10): 4729-47 32. Brown, D. M., Upcroft, J. A. and Upcroft, P. (1996). A H2O producing NADH oxidase from the protozoan parasite Giardia duodenalis. Eur J Biochem. 241:155-l6L Bsat, N., Herbig, A'., Casillasmartinez, L., Setlow, P. and Helmann, J. D. (1998). Bacillus subtilis contains multiple fur homologues - identification of the iron uptake (fur) and peroxide regulon (perR) r'epressors. Mol Microbiol. 29(I): 1 89- 1 98. Buckel, W. (2001). Unusual enzymes involved in five pathways of glutamate fermentation. Appl Microbiol Biotechnol . 5l (3): 263 -21 3 . Caldwell, C. E. and Marquis, R. E. (1999). Oxygen metabolism by Treponema denticola. Oral Microbiol Immunol. 14: 66-12. Candeias, L. P., Patel, K. 8., Stratford, M. R. L. and Wardman' P. (1993). Free hydroxyl radicals are formed on reaction between the neutrophil- 213 derived species superoxide anion and hypochlorous acid. FEBS Lett. 333(t-2): 151-153. Carlsson, J. W., Höfling, J. F. and Sundqvist, G. K. (1984). Degradation of albumin, haemopexin, haptoglobin and transferrin by black-pigmented Bacteroides species. J Med Microbiol. 18:39-46. Carmel-Harel, O. and Storz, G. (2000). Roles of the glutathione- and thioredoxin-dependent reduction systems in the Escherichia coli and Saccharomyces cerevisiae responses to oxidative stress. Ann Rev Microbiol. 54: 439-461. Chen, W. and Kuo, T. (1993). A simple and rapid method for the preparation of Gram-negative bacterial genomic DNA. Nucleic Acids Res. 21(9): 2260. Costerton, J. W., Cheng, K. J. and Geesey, G. G. (1987). Bacterial biofilms in nature and disease. Ann Rev Microbiol. 69: 436-44L Dashper, S. G., Brownfield, L., Slakeski, N., Zilm, P. S., Rogers, A. H. and Reynolds, E. C. (2001). Sodium ion-driven serine/threonine transport in P orphyromonas girtgivalis. J Bacteriol. 1 83( 14) : 4142-4148. DiRita, V. J. and Mekalanos, J. J. (1989). Genetic regulation of bacterial virulence. Ann Rev Genet. 23:455-482. Ebersole, J. L., Feuille, F., Kesavalu, L. and Holt, S. C. (1997). Host modulation of tissue destruction caused by periodontopathogens - effects on a mixed microbial infection composed of Porphyromonas gingivalis and Fusobctcîerium nucleatum. Microb Pathogen . 23(I): 23-32' Farquharson, S.I., Germaine, G.R., Gray, G.R. (2000)' Isolation and characterization of the cell-surface polysaccharides of Porplryromonas gingivalis ATCC 53918. Oral Microbiol Immunol. 15(3): I5I-151 . 214 Fedi, P. F. and Killoy, W. J. (1992). Temperature differences at periodontal sites in health and disease. J Periodontol. 63: 24-21. Fitip, C., Fletcher, G., Wulff, J. L. and Earhart, C. F. (1973). Solubilization of the cytoplasmic membrane of Escherichia coli by the ionic detergent sodium lauryl sarcosinate. J Bacteriol. 1 15(3): 7ll -122. Fletcher, H. M., Morgan, R. M. and Macrina, F. L. (1991). Nucleotide sequence of the Porphyromonas gingivalis W83 recA homolog and construction of a recA-deficient mutant. Infect lmmun. 65(11): 4592-4597. Fletcher, J., Nair, S., Poole, S., Henderson, B. and Wilson, M. (1998). Cytokine degradation by biofilms of Porphyromonas gingivalis. Curr Microbiol. 36(4):216-2t9. Forng, R.-Y., Champagne, C., Simpson, W. and Genco, C. A. (2000). Environmental cues and gene expression in Porphyromonas gingivalis and Ac t inob ac illus act inomy c e t e mc omit an s . Oral Diseases. 6 : 3 5 1 - 3 65. Fukuda, T., Takenaka, S., Iwaku, M. and Hoshino, E. (2001). Bacterial aerotolerance assay with an artificial biofilm model. J Dent Res. 8O(Special Issue -IADR Abstracts): 575. Genco, C. A. and Dixon, D. W. (2001). Emerging strategies in microbial haem capture. Mol Microbiol. 39(1): 1-11. Gerhardt, 4., Cinkaya,1., Linder, D., Huisman, G. and Buckel, W. (2000). Fermentation of 4-aminobutyrate by Clostridium aminobutyricum: cloning of two genes involved in the formation and dehydration of 4- hydroxybutyryl-CoA. Arch Microbiol. 17 4(3): 1 89- 1 99. 215 Gerlach, D., Reichardt, W. and Vettermann, S. (1998). Extracellular superoxide dismutase from Streptococcus pyogetxs type 12 strain is manganese- dependent. FEMS Microbiol Lett. 160: 2ll-224. Gharbia, S. E. and Shah, H. N. (1991). Utilization of aspartate, glutamate and their corresponding peptides by Fusobacterium nucleatutn subspecies and Porphyromonas gittgivalis. Cwr Microbiol. 22: 159-163. Gharbia, S. 8., Shah, H. N. and Welch, S. G. (1989). The influence of peptides on the uptake of amino acids in Fusobacterium: predicted interaction with Porphyronxonas gingivalis. Curu Microbiol. 19: 23I-235. Gibbons, R. J. (1964). Bacteriology of dental caries. J Dent Res. 43: 102l-1028' Gibello, 4., Suarez., M., Allende, J. L. and Martin, M. (1991). Molecular cloning and analysis of the genes encoding the 4-hydroxyphenylacetate hydroxylase from Klebsiella pneumoniae. Arch Microbiol. 161(2-3): 160- 166. Gibson, C. M., Mallett, T. C., Claiborne, A. and Caparon, M. G. (2000). Contribution of NADH oxidase to aerobic metabolism of Streptococcus pyogenes. J Bacteriol. 182(2): 448-455. Gish, W . (199 6-2000). BLASTP 2.0. http ://blast.wustl.edu. Gmür, R., Hrodek, K., Saxer, U. P. and Guggenheim, B. (1986). Double blind analysis of the relation between adult periodontitis and systemic host response to suspected periodontal pathogens. Infect Immun. 52:168-116. Gomes, C. M. and Texeira, M. (1998). The NADH oxidase from the thermoacidophilic archaea Acidianus ambivalens: isolation and physicochemical characterisation. Biochem Biophys Res Commun. 243 (2): 412-415. 216 Gonzalez-Flecha, B. and Demple, B. (1991). Homeostatic regulation of intracellular hydrogen peroxide concentration in aerobically growing Escherichia coli. J Bacteriol . Il9(2): 382-388. Grenier, D., Mayrand, D. and McBride, B. (1989). Further studies on the degradation of immunoglobulins by black-pigmented Bacteroides. Oral Microbiol Immunol. 4: 12-18. Guerrant, G. O., Lambert, M. A. and Moss, C. W. (1982). Analysis of short- chain fatty acids from anaerobic bacteria by high-performance liquid chromathography. J Clin Microbiol. 16: 355-360. Hahn, J.-S. and Roe, J.-H. (1999). OxyR sequence ftom Streptomyces coelicolor A3 (2). Direct submis sion to NCB I. http : //www. ncbi'nlm. nih. gov. Harms, C., Meyer, M. A. and Andreesen, J. R. (1998). Fast purification of thioredoxin reductases and of thioredoxins with an unusual redox-active centre from anaerobic, amino-acid-utilizing bacteria. Microbiol. 144(Pan 3):793-800. Harris, E.D. (1992). Regulation of antioxidant enzymes. FASEB J.6:2675-2682. Higuchi, M. (1934). The effect of oxygen on the glowth and mannitol fermentatio n of Streptococcus mutants . J Gen Microbiol. 130: 1 8 19-1826. Higuchi, M. (1992). Reduced nicotinamide adenine dinucleotide oxidase involvement in defense against oxygen toxicity of Streptococcus mutans. Oral Microbiol Immunol. 7(5): 309-14. Higuchi, M., Shimada, M., Matsumoto, J., Yamamoto, Y., Rhaman, A. and Kamio, Y. (1994). Molecular cloning and sequence analysis of the gene encoding the H2O2-forming NADH oxidase from Streptococcus mutans. Bioscience Biotechnol Biochem, 58: 1603-1607. 217 Higuchi, M., Shimada, M., Yamamoto, Y., Hayashi, T., Koga, T. and Kamio, Y. (1993). Identification of two distinct NADH oxidases coresponding to HzOz-forming oxidase and HzO-forming oxidase induced in Streptococcus mutans. J Gen Microbiol. I39:2343-2351. Hojrup, P. General Protein Mass Analysis for Windows (GPMAV/). Odense, Denmark. Lighthouse data. Holt, S. C., Kesavalu, L., Walker, s. and Genco, c. A. (1999). Virulence factors of Porphyromonas gingivalis. Periodontol 2000. 20 168-238. Ilouno, L. 8., Shu, E. N. and lgbokwe, G. E. (1996). An improved technique for the assay of red blood cell superoxide dismutase (SOD) activity. Clin Chim Acta.24'7: l-6. Imlay, J. A. (1995). A metabolic enzyme that rapidly produces superoxide, fumarate reductase of Escherichia coli. J Biol Chem. 210(34): 19767- 19711. Imlay, J. A. (2002). How oxygen damages microbes: Oxygen tolerance and obligate anaerobiosis. Adv Microb Physiol. 46(46): 111-153. Iverson, T. M., Luna-Chavez, C., Cecchini, G. and Rees, D. C. (1999). Structure of the Escherichia coli fumarate reductase respiratory complex. Science. 284(5422): 196l - 1966. Jackson, C. 4., Hoffmann, 8., Slakeski, N., Cleal, S., Hendtlass' A. J. and Reynolds, E. C. (2000). A consensus Porphyromonas gingivalis promoter sequence. FEMS Microbiol Lett, 186(1): 133-138. Jakubovics, N. S., Smith, A. W. and Jenkinson' H. F. (2002)' Oxidative stress tolerance is manganese (Mn'*) regulated in Streplococcus gordonii. Microbiol. 148 3255-3263. 218 Jenney, F. E., Verhagen, M., Cui, X. Y. and Adams, M. W. W. (1999). Anaerobic microbes: Oxygen detoxification without superoxide dismutase. Science. 286(5438): 306-309. Kadowaki, T., Yoneda, M., Okamoto, K., Maeda, K. and Yamamoto, K. (1994). Purification and characterization of a novel arginine-specific cysteine proteinase (argingipain) involved in the pathogenesis of periodontal disease from the culture supernatant of Porphyromonas gingiv alis. J Biol Chem. 269 (33): 2131 I -2131 8' Kamaguchi,4., Nakayama, K., OhyamãrT., Watanabe, T., Okamoto, M. and Baba, H. (2001). Coaggregation of Porphyromonas gittgivalis and P revot ella intermed ia. Microbiol Immunol. 45 (9): 649 -656. Kapatral, V., Anderson, I., Ivanova, N., Reznik, G., Los, T., Lykidis, 4., Bhattacharyya, 4., Bartman, 4., Gardner, W., Grechkin, G., Zhu, L. H., Vasieva, O., Chu, L., Kogan, Y., Chaga, O., Goltsmanr E.' Bernal, 4., Larsen, N., D'Sottza, M, Walunas, T., Pusch, G., Haselkorn, R., Fonstein, M., Kyrpides, N. and Overbeek, R. (2002)' Genome sequence and analysis of the oral bacterium Fusobacterium nucleatum strain ATCC 25586. J Bacteriol. 184(7): 2005-2018. Kesavalu, L., Holt, S. C. and Ebersole, J. L. (1996)' Trypsin-like protease activity of Porphyromonas gingivalis as a potential virulence factor in a murine lesion model. Microb Pathogen. 20(1): 1-10. Kinane, D. F. (2001). Causation and pathogenesis of periodontal disease. Periodontol 2000. 25: 8-20. 219 Kinane, D. F., Mooney, J. and Ebersole, J.L. (1999). Humoral immune response Io Actinob acill us actinomyc etemc omit atts and P orphyromonas gin giv alis in periodontal disease. Periodontol 2000. 20: 289 -3 40. Klenk, H. P., Clayton, R. 4., Tomb, J.-F., White, O., Nelson, K.E., Ketchum, K. 4., Dodson, R. J., Gwinn, M., HickeY, E. K., Peterson, J. D.' Richardson, D. L., Kerlavage, A. R., Graham, D.8., Kyrpides, N. C., Fleischmann, R. D., Quackenbush, J., Lee, N. H., Sutton, G. G., Gill, S., Kirkness, E. F., Dougherty, B. 4., McKenney, K., Adams, M. D., Loftus, 8., Peterson, S., Reich, C. I., McNeil, L. K., Badger, J. H., Glodek, A,., Zhou, L., Overbeek, R., Gocayne, J. D., Weidman, J. F.' McDonald, L., Utterback, T., Cotton, M. D., Spriggs, T., Artiach, P., Kaine, B. P., Sykes, S. M., Sadow, P. W., D'Andrea, K. P., Bowman, C., Fujii, C., Garland, S. 4., Mason, T. M., Olsen, G. J., Fraser, C. M., Smith, H. O., Woese, C. R. and Venter, J. C. (1991). The complete genome sequence of the hyperthermophilic, sulphate-reducing archaeon Archaeoglobus fulgid¿¿s. Nature. 390(6658): 364-310. Kolenbrander, P. E., Andersen, R. N. and Moore, L. V. (1989). Coaggregation of Fusobacterium nucleatum, Selenomottas flueggei, Selenomonas infelix, Selenomonas noxia, and Selenomotxas sputigena with strains from 11 genera of oral bacteria. Infect Immun. 57(10): 3194-203. Kolenbrander, P. E., Ganeshkumar, N., Cassels, F. J. and Hughes, C. V. (1993). Coaggregation: specific adherence among human oral plaque bacteria. FASEB I. 7 : 406-413. 220 Kullik,I., Stevens, J., Toledano, M. B. and Storz, G. (1995). Mutational analysis of the redox-sensitive transcriptional regulator oxyR - regions important for DNA binding and multimerization. J Bacteriol. I17(5): 1285-1291. Kurita-Ochiai, T., Ochiai, K. and Fukushima, K. (2000). Butyric-acid-induced apoptosis in murine thymocytes and splenic T- and B-cells occurs in the absence of p53. J Dent Res. 79(12): 1948-1954. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacterioph age T 4. Nature. 221 : 680-685 . Laine, M. L., Appelmelk, B.J. and van Winkelhoff' A.J. (1991). Prevalence and distribution of six capsular serotypes of Porphyromonas gingivalis in periodontitis patients. J Dent Res. 76(12): 1840-1844. Laine, M. L. and van Winkelhoff, A. J. (1998). Virulence of six capsular serotypes of Porphyromotxas girtgivalis in a mouse model. Oral Microbiol Immunol. 13(5): 322-325. Lamont, R. J., Chan,4., Belton, C.M.,Izutsu, K. T., Vasel, D. and Weinberg, A. (1995). Porphyromonas gingivalis invasion of gingival epithelial cells. Infect Immun. 63(10): 3878-3885. Lamont, R. J., El-Sabaeny, 4., Park, Y., Cook, G. S., Costerton, J. W. and Demuth, D. R. (2002). Role of the Streptococcus gordonii SspB protein in the development of Porphyron'Lonas gingivalis biofilms on streptococcal substrates. Microbiol. 148(Pafi 6): 1621 -1636. Lamont, R. J. and Jenkinson, H. F. (1998). Life below the gum line: pathogenic mechanisms of Porpltyromonas gingivalis. Microbiol Mol Biol Rev. 62(4): t244-63. 22t Leke, N., Grenier, D., Goldner, M. and Mayrand, D. (1999). Effects of hydrogen peroxide on growth and selected properties of Porphyromonas gingiv alis. FEMS Microbiol Lett. 11 4(2) : 341 -3 53. Liu, Y. and Fletcher, H. M. (2001). Environmental regulation of recA gene expression in Porphyromorxas gingivalis. Oral Microbiol Immunol. 16(3): 136-143. Liu, Y. and Fletcher, H. M. (2001). The recA gene in Porphyromonas gingivalis is expressed during infection of the murine host. Oral Microbiol Immunol. 16(4):218-223. Loewen, P. C. (1996). Escherichia coli PerR (perR) gene, complete cds. Direct submission to NCBI. http://www.ncbi.nlm.nih.gov. Lopez de Felipe, F. and Hugenholtz, J. (1999). Pyruvate flux distribution in NADH-oxidase-overproducing Lactococcus lactis strain as a function of culture conditions. FEMS Microbiol Lett. 119: 46I-466. Lourbakos, 4., Yuan, Y. P., Jenkins, A. L., Travis, J., Andrade-Gordon, P., Santulli, R., PotemPa, J. and Pike, R. N. (2001)' Activation of protease- activated receptors by gingipains from Porphyromonas gingivølls leads to platelet aggregation: a new trait in microbial pathogenicity. Blood. 9l(12): 3190-3197. Lynch, M. C. and Kuramitsu, H. K. (1999). Role of superoxide dismutase activity in the physiology of Porphyromonas gingivalis. Infect Immun. 67(1): 3361-331s. Maciver, I. and Hansen, E. J. (1996). Lack of expression of the global regulator OxyR in Haemophilus iffiuenzae has a profound effect on growth phenotype. Infect Immun. 64(11): 4618-4629. 222 Madianos, P. N., Papapanou, P. N., Nannmark, U. and Dahlfen, G. (1996). Porphyromonas gingivalis FDC381 multiplies and persists within human oral epithelial cells in vitro.Infect Immun. 64(2): 660-664. Madigan, M. T., Martinko, J. M. and Parker, J. (1991). Brock: Biology of Microorganisms. New Jersey, Prentice Hall International, Inc. Maley, J., Shoemaker, N. and Roberts, I. S. (1992). The introduction of colonic- Bacteroides shuttle plasmids info Porphyromonas gingivalis: Identification of a putative P. gingivalis insertion-sequence element. FEMS Microbiol Lett.93: 15-82. Marsh, P. D., Mc Dermid, A. S., McKee, A. S. and Baskerville, A'. (1994). The effect of growth rate and haemin on the virulence and proteolytic activity of Porphyromonas gingivalis W50. Microbiol. 140: 861-865. Marsh, P.D. and Bradshaw, D.J. (1997). Physiological approaches to the control of oral biofilms. Adv Dent Res. 11(1): 176-185. Martins, E. A. L. and Demple, B. (1996). Salmottella typhimurium oxidative stress regulon, SoxS (saxS) and SoxR (soxR) genes, complete cds. Direct submission to NCBI. http://www.ncbi.nlm.nih.gov' Massey, V. (1994). Activation of molecular oxygen by flavins and flavoproteins. J Biol Chem. 36: 22459-22462. Masuda, K., Tomita, K., Hayashi, H., Yoshioka, M., Hinode, D. and Nakamura, R. (2001). Consumption of peptide-derived arginine by a periodontopathogenic bacterium, Porphyromonas gingivalis. Anaerobe. 1(4):209-211. 223 Masuda, K., Yoshioka, M., Hinode, D. and Nakamura' R. (2002). Purification and characterization of arginine carboxypeptidase produced by P o rp hy r omonas g irt g iv ali s . Infect Immun. 1 O(4) : I 807 - 1 8 1 5. Matsumoto, J., Higuchi, M., Shimada, M., Yamamoto, Y. and Kamio, Y. (1996). Molecular cloning and sequence analysis of the gene encoding the HzO-forming NADH oxidase from Streptococcus mutans. Biosc Biotechnol Biochem.60(1): 39-43. Mayrand, D. and Holt, S. C. (1988). Biology of asaccharolytic black-pigmented Bacteroides species. Microbiol Rev. 52: 134-152. McCord, J. M. and Fridovich, I. (1969). Superoxide dismutase' An enzymic function for erythrocuprein (hemocuprein). J Biol Chem.244(22): 6049-55. McDermid, A. S., McKee, A. S. and Marsh, P. D. (1988)' Effect of environmental pH on enzyme activity and growth of Bacteroides gingivalis W50. Infect Immun. 56: 1096-1100. McKee, A. S., McDermid, A. S., Baskervill€,4., Dowsett, A.8., Ellwood, D. C. and Marsh, P. D. (1986). Effect of hemin on the physiology and virulence of Bacteroides gingivalls W50. Infect lmmun. 52(2):349-355. Meister, L. (1992). A train of research: from glutamine synthetase to selective inhibition of glutathione synthesis. Chemtracts Biochem Mol Biol. 3: 75- 106. Messner, K. R. and Imlay, J. A. (1999). The identification of primary sites of superoxide and hydrogen peroxide formation in the aerobic respiratory chain and sulfite reductase complex of Escherichia coli. J Biol Chem. 274(t5): 101 19-10128. 224 Miller, R. A. and Britigan, B. E. (1997). Role of oxidants in microbial pathophysiology. Clin Microbiol Rev. 10(1). Mongkolsuk, S. and Helmann, J. D. (2002). Regulation of inducible peroxide stress responses. Mol Microbiot. 45(1): 9-15. Mongkolsuk, S., Praituan, W., Loprasert, S., Fuangthong, M. and Chamnongpol, S. (1998). Identification and charactertzation of a new organic hydroperoxide resistance (ohr) gene with a novel pattern of oxidative stress regulation ftom Xanthomonas campestris pv. phaseoli. J Bacreriol. 180(10): 2636-2643. Mooney, J. and Kinane, D. F. (1994). Comparison of humoral immune response to Porphyromonas gingivalis and Actinobacillus actinomycetemcomitans in adult periodontitis and rapidly progressive periodontitis. Oral Microbiol Immunol. 9:32I-326. Moore, W. E. C. and Moore, L. V. H. (1994). The bacteria of periodontal diseases. Periodontol 2000. 5: 66-71. Morgan, R. W., Christman, M. F., Jacobson, F. S., Storz, G. and Ames, B. N. ( 1 986). Hydrogen peroxide-inducible proteins in S almonella thy phimurium overlap with heat shock and other stress proteins. Proc Natl Acad Sci USA. 83: 8059-8063. Mouton, C., Hammond, P. G., Slots, J. and Genco, R. J. (1981). Serum antibodies to oral Bacteroides asaccharolyticus (Bacteroides girtgivalis): relationship to age and periodontal disease. Infect Immun. 3I: 182-192. Müh, U., Buckel, W. and Bill, E. (1991). Mössbauer study of 4-hydroxybutyryl- CoA dehydratase. Probing the role of an iron-sulfur cluster in an overall non-redox reaction. Eur J Biochem. 248: 380-384. 225 Nassar, H., Chou, H. H., Khlgatian, M., Gibson, F. C., Van Dyke, T. E. and Genco, C. A. (2002). Role for fimbriae and lysine-specific cysteine proteinase gingipain K in expression of interleukin-8 and monocyte chemoattractant protein in Porphyromonas gingivalis-infected endothelial cells. Infect Immun. 70(1): 268-216. Neidhardt, F. C., Ingraham, J. L. and Schaechter, M. (1990). Physiology of the bacterial cell. A molecular approach. Sunderland, Massachusetts, Sinauer Associates,Inc. Nguyen, P. T. M., Abranches, J., Phan, T. N. and Marquis, R. E. (2002). Repressed respiration of oral streptococci grown in biofilms. Curr Microbiol. 44(4): 262-266. Niimura, Y., Nishiyama, Y., Saito, D., Tsuji, H., Hidakâ, M., Miyaji' T.' Watanabe, T. and Massey, V. (2000). A hydrogen peroxide-forming NADH oxidase that functions as an alkyl hydroperoxide reductase in Amphibacillus rylanus. J Bacteriol. 182(18): 5046-5051. Niimura, Y., Ohnishi, K., Yarita, Y., Hidaka, M., Masaki, H., Uchimura, T., Suzuki, H., Kozaki, M. and Uozumi, T. (1993). A flavoprotein functional as NADH oxidase ftom Amphibacillus xylanus ep01 - purification and characterization of the enzyme and sffuctural analysis of its gene. J Bacteriol. I1 5(24): I 945-1 950. Niimura, Y., Poole, L. B. and Massey, V. (1995). Amphibacillus xltlanus NADH oxidase and Salmonella typhimurium alkyl-hydroperoxide reductase flavoprotein components show extremely high scavenging activity for both alkyl hydroperoxide and hydrogen peroxide in the presence of S. 226 typhimurium alkyl-hydroperoxide reductase 22-kDa protein component. J Biol Chem. 2lO(43): 25645-25650. Nishiyama, Y., Massey, V., Takeda, K., Kawasaki, S., Sato, J., Watanabe, T. and Niimura, Y. (2001). Hydrogen peroxide-forming NADH oxidase belonging to the peroxiredoxin oxidoreductase family: Existence and physiological role in bacteria. J Bacteriol. 183(8): 2431-2438. O'Brien-Simpson, N. M., Paolini, R.4., Hoffman,8., Slakeski, N., Dashper, S. G. and Reynolds, E. C. (2001). Role of RgpA, RgpB, and Kgp proteinases in virulence of Porphyromonas gingivalis W50 in a murine lesion model. Infect Immun. 69(12): 1521-1534. Okamoto, K., Nakayama, K., Kadowaki, T., Abe, N., Ratnayake, D. B. and Yamamoto, K. (1998). Involvement of a lysine-specific cysteine proteinase in hemoglobin adsorption and heme accumulation by P orphyromonas g irt givalis. J Biol Chem . 27 3 (33): 21225 -2123 l. Oliver, R. C., Brown, L. J. and Loe, H. (1989). An estimate of periodontal treatment needs in the U.S. based on epidemiological data. J Periodontol. 60: 371-380. Oliver, R. C., Brown, L. J. and Loe, H. (1998). Periodontal diseases in the United States population. J Periodontol. 69 (2) : 269 -27 8. Park, M. K., Myers, R. A. and Marzella, L. (1992). Oxygen tensions and infections: modulation of microbial growth, activity of antimicrobial agents, and immunologic responses. Clin Infect Dis. 14(3): 720-40. Parkhill, J., Dougan, G., James, K. D., Thomson, N. R., Pickard, D., Wain, J., Churcher, C., Mungall, K. L., BentleY, S. D., Holden, M. T. G., Scbaihia, M., Baker, S., Basham, D., Brooks, K., Chillingworth, T., 221 Connerton, P., Cronin, A.., Davis, P., Davies, R.M., Dowd, L., White, N., Farrar, J., Feltwell, T., Hamlin, N., Haque,A'., Hien, T. T., Holroyd, S., Jagels, K., Krogh,4., Larsen, T. S., Leather, S., Moule, S., O'Gaora, P., Parry, C., Quail, M., Rutherford, K., Simmonds, M., Skelton, J., Stevens, K., Whitehead, S. and Barrell, B. G. (2001). Complete genome sequence of a multiple drug resistant Salmonella enterica serovar Typhi CT18. Nature. a13(6858): 848-852. Parkhill, J., Wren, B. W., Thomson, N. R., Titball, R. W., Holden, M. T. G.' Prentice, M.8., Sebaihia, M., James, K. D., Churcher, C., Mungall, K. L., Baker, S., Basham, D., BentleY, S.D., Brooks, K., Cerdeno'Tarraga, A. M., Chillingworth, T., Cronin,4., Davies, R. M., Davis, P., Dougan, G., Feltwell, T., Hamlin, N., Holroyd, S., Jagels, K., Karlyshev, A. V. and et al. (2001). Genome sequence of Yersinia pestis, the causative agent of plague. Nature. 413(6855): 523-521 . Percival, R. S., Marsh, P. D., Devine, D.4., Rangarajatr, M., Aduse-Opoku, J., Shepherd, P. and Curtis, M. A. (1999). Effect of temperature on growth, hemagglutination, and protease activity of Porphyromonas girtgivalis. lnfect Immun. 61(4): 1917-1921. Pike, R. N., Potempa, J., McGraw, W., Coetzer, T. H. T. and Travis, J. (1996)' Characterization of the binding activities of proteinase-adhesin complexes ftom Porphyromonas gingivalis. J Bacteriol. i78(10): 2816-2882' Pollack, J. D., Williams, M. V. and McElhaney, R. N. (1997). The comparative metabolism of the mollicutes (mycoplasmas): the utility for taxonomic classification and the relationship of putative gene annotation and 228 phylogeny to enzymatic function in the samllest free-living cells. Crit Rev Microbiol. 23: 269-354. Pomposiello, P. J., Bennik, M. H. J. and Demple, B. (2001). Genome-wide transcriptional profiling of the Escherichia coli responses to superoxide stress and sodium salicylate. J Bacteriol. 183(13): 3890-3902. Pomposiello, P. J. and Demple, B. (2002). Global adjustment of microbial physiology during free radical stress. Adv Microb Physiol. 46(46): 319- 34r. Poole, L. B. and Claiborne, A. (1986). Interactions of pyridine nucleotides with redox forms of the flavin containing NADH peroxidase from Streptococcus faecalis. J Biol Chem. 261: 14525-14533. Poole, L. 8., Higuchi, M., Shimada, M., Calzi, M. L. and Kamio, Y. (2000). Streptococcus mutans H2O2-forming NADH oxidase is aî alkyl hydroperoxide reductase protein. Free Radic Biol Med. 28(1): 108-20. Poyart, C., Quesn€, G., Coulon, S., Berche, P. and Trieu-Cuot, P. (1998). Identification of streptococci to species level by sequencing the gene encoding the manganese-dependent superoxide dismutase. J Clin Microbiol. 36:41-47. Prieto, M. A. and Garcia, J. L. (1994). Molecular characterization of 4- hydroxyphenylacetate 3-hydroxylase of Escherichia coli. A two-protein component enzyme. J Biol Chem. 269(36):22823-22829. Rangarajan, M., Smith, S. J. M. and Curtis, M. A. (1991). Biochemical characterization of the arginine-specific proteases of Porphyromonas gingivalis \M50 suggests a common precursor. Biochem J . 323(Part 3): l0I- 709. 229 Reilly, S. (1980). The carbon dioxide requirements of anaerobic bacteria. J Med Microbiol. 13: 51 3-57 9. Rlizza,V., Sinclair, P. R., D.C., W. and Cuorant, P. R. (1968). Electron transport system of the protoheme requiring anaerobe Bacteroides melaninogenicus. J Bacteriol. 96(3): 665-611. Robrish, S. 4., Fales, H. M., Gentry-Weeks, C. and Thompsom' J. (1994)' Phosphoenolpyruvate-dependent maltose: phosphotransferase activity in Fusobacterium mortiferum ATCC 25551: specificity, inducibility and product analysis. J Bacteriol . I1 6(3250-325 6). Rocha, E. R., Owens, G. and Smith, c. J. (2000). The redox-sensitive transcriptional activator OxyR regulates the peroxide response regulon in the obligate anaerobe Bacteroides fragilis. J Bacteriol. 182(18): 5059-5069. Rocha, E. R., Selby, T., Coleman, J. P. and Smith, C. J. (1996). Oxidative stless response in an anaerobe, Bacteroides fragilis - a role for catalase in protection against hydrogen peroxide. J Bacteriol . 118(23):6895-6903. Rocha, E. R. and Smith, C. J. (1993). Characterization of a peroxide-resistant mutant of the anaerobic bacterium Bacteroides fragilis. J Bacteriol. I8O(22): 5906-5912. Rocha, E. R. and Smith, C. J. (1999). Role of the alkyl hydroperoxide reductase (ahpCÐ gene in oxidative stress defense of the obligate anaerobe Bacteroides fragilis. J Bacteriol. 181(18): 5701-5710. Rogers, A. H., Gunadi, 4., Gully, N. J. and Zilm, P. S. (1998). An aminopeptidase nutritionally important to Fusobacterium nucleatunt. Microbiol. 144(Part 7): 1 807- 1 8 13. 230 Ross, R. P. and Claiborne, L. (1992). Molecular cloning and analysis of the gene encoding the NADH oxidase from Streptococcus faecalis 10C1. Comparison with NADH peroxidase and the flavoprotein disulfide reductases. J Mol BioI. 221 : 658-61 I. Sabra, W., Kim, E.-J. and Zeng, A.-P. (2002). Physiological responses of Pseudomonas aeruginosa PAOl to oxidative stress in controlled microaerobic and aerobic cultures. Microbiol. I48: 3 19 5 -3202. Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Moiecular Cloning. A laboratory manual., Cold Spring Harbor Laboratory Press. Sandros, J., Karlssotr, C., Lappin, D. F., Madianos, P. N., Kinane, D. F. and Papapanou, P. N. (2000). Cytokine responses of oral epithelial cells to Porphyromonas gingivølis infection. J Dent Res. 79(10): 1808-1814, Scannapieco, F. 4., Millar, S. J., Reynolds, H. S., Zambon, J. J. and Levine, M. J. (1987). Effect of anaerobiosis on the surface ultrastructure and surface proteins of Actinobacillus actinomycetemcomitatts (Haemoplúlus actinomyc etemc omitans). Infect Immun. 55 (9) : 2320-2323 . Scherf, U. and Buckel, W. (1993). Purification and properties of an iron-sulfur and FAD-containing 4-hydroxybutyryl-Coa dehydratase vinylacetyl-Coa delta(3)-delta(2)-isomerase from Clostridium aminobutyricum. Eur J Biochem. 215(2): 421-429. Schifferle, R. E., Reddy, M. S., Zambon, J. J., Genco, R. J. and Levine M. J. (1989). Characterization of a polysaccharide antigen ftom Bacteroides gittgivalis. J Immunol . I43: 3035-3042. Schubert, J. and Wilmer, J. W. (1991). Does hydrogen peroxide exist "free" in biological systems? Free Radic Biol Med. 11: 545-555. 231 Shah, H. N. and Collins, M. D. (1988). Proposal for the reclassification of Bacteroides asaccharolyticus, Bacteroides girtgivalis, and Bacteroides endodontalis in a new genus , Porphyromonas.Int J Syst Bacteriol. 38: 128- 131. Shah, H. N. and Gharbia, S. E. (1989). Ecological events in subgingival dental plaque with reference to Bacteroides and Fusobacterium species. lnfection. Il:264-268. Shah, H. N., Williams, R. 4., Bowden, G. H. and Hardie, J. M. (1976). Comparison of the biochemical properties of Bacteroides melaninogenicus from human dental plaque and other sites. J Appl Bacteriol. 41(3): 473-95. Shelburne, C. E., Gleason, R. M., Germaine, G. R., Wolff, L. F., Mullally, B. H., Coulter, W. A. and Lopatin, D. E. (2002). Quantitative reverse transcription polymerase chain reaction analysis of Porphyromonas gingivalis gene expression iz vivo. J Microbiol Meth. 49(2): l4l-156. Shi, Y., Ratnayake, D. 8., Okamoto, K., Abe, N., Yamamoto, K. and Nakayama, K. (1999). Genetic analyses of proteolysis, hemoglobin binding, and hemagglutination of Porphyromonas girtgivalis - Construction of mutants with a combination of rgpA, rgpB, kgp, and hagA. J Biol Chem. 274(25): 17955-11960. Shyu, Y. C. and Lin, F. P. (1999). Cloning and characterization of manganese superoxide dismutase gene from Vibrio parahaemolyticus and application to preliminary identification of Vibrio strains. Iubmb Life. 48(3): 345-352. Sies, H., Ed. (1985). Oxidative stress. London, Academic Press. Sigler, K., Chaloupka, J., Brozmanova, J., Stadler, N. and Höfer, M. (1999). Oxidative stress in microorganisms-L - I - Microbial vs. higher cells - 232 Damage and defenses in relation to cell aging and death. Folia Microbiolo gica. 44(6): 587 -624. Slakeski, N., Bhogal, P. S., O'Brien-Simpson, N. M. and Reynolds, E. C. (1998). Cbaracteúzation of a second cell-associated Arg-specific cysteine proteinase of Porphyromonas gingivalis and identification of an adhesin- binding motif involved in association of the prtR and prtK proteinases and adhesins into large complexes. Microbioì. 144(Part 6): 1583-1592. Slots, J. and Chen, C. (1999). The oral microflora and human periodontal disease. In: Medical importance of the normal microflora. G. W. Tannock. Great Britain, Kluwer Academic Publishers: l0l-121. Smalley, J. W., Birss, A. J., McKee, A. S. and Marsh, P. D. (1991). Haemin- restriction influences haemin binding, hemagglutination and protease activity of cells and extracellular membrane vesicles of Poprhyromonas girtgivalis W50. FEMS Microbiol Lett. 90(63-68). Smalley, J. W., Birss, A. J. and Silver, J. (2000). The periodontal pathogen Porphyromonas gingivalis harnesses the chemistry of the micro-oxo bishaem of iron protoporphyrin IX to protect against hydrogen peroxide. FEMS Microbiol Lett. 183(1): 159-164. Smalley, J. W., Silver, J., Marsh, P. D. and Birss, A. J. (1998). The periodontopathogen Porphyromonas gingiv¿lls binds iron protoporphyrin D( in the mu-oxo dimeric form: an oxidative buffer and possible pathogenic mechanism. Biochem J. 331(Pt 3): 681-5. Smith, M. A. and Edwards, D. l. (1991). Oxygen scavenging NADH oxidase and metronidazole resistance in Helicobacter pylori. J Antimicrob Chemother. 39:347-353. 233 Socransky, S. S. and Haffajee, A. D. (1994). Evidence of bacterial etiology: a historical perspective. Periodontol 2000. 5: 7 -25. Socransky, S. S., Haffajee, A. D., Cugini, M.4., Smith, C. and Kent, R. L., J.. (1998). Microbial complexes in subgingival plaque. J Clin Periodontol. 25(2): 134-44. Stadtman, E. R. (1992). Protein oxidation and aging. Science. 257: 1220-1224. Stanton, T. (1989). Glucose metabolism and NADH recycling by Treponema hyodysenteriae, the agent of swine dysentery. Appl Environ Microbiol. 55: 2365-231r. Stanton, T. B. and Jensen, N. S. (1993). Purification and characterization of NADH oxidase from Brachyspira (Serpulina) hyodysenteriae. J Bacteriol. 175:2980-2981. Stanton, T. 8., Rosey, E. L., Kennedy, M. J., Jensen, N. S. and Bosworth, B. T. (1999). Isolation, oxygen sensitivity and virulence of NADH oxidase mutants of the anaerobic spirochete Brachyspira (Serpulina) ltyodT,s¿n¡sr'¡6¿, etiologic agent of swine dysentery. Appl Environ Microbiol. 65: 5028-5034. Storz, G., Tartaglia, L. 4., Farr, S. B. and Ames, B. N. (1990). Bacterial defences against oxidative stress. Trends Genet. 6:363-368. Storz, G. and Zheng, M. (2000). Oxidative stress. In: Bacterial stress responses. Ed. Storz, G. and Hengge-Aronis, R, Washington, DC, ASM Press: 4l-59. Syed, S. 4., Makinen, K. K., Makinen, P. L., Chen, C. Y. and Muhammad,, Z. (1993). Proteolytic and oxidoreductase activity of Treponema-denticola ATCC 35405 grown in an aerobic and anaerobic gaseous environment. Res Microbiol. 144(4): 311 -326. 234 Sztukowskâ, M., Bugno, M., Potempa, J., Travis, J. and Kurtz, D. M. (2002). Role of rubrerythrin in the oxidative stress response of Porphyromonas gingivalis. Mol Microbiol. 44(2): 41 9 -488. Tada, H., Sugawara, S., Nemoto, E., Takahashi, N.r Imamura, T., Potempa, J., Travis, J., Shimauchi, H. and TakadarH. (2002). Proteolysis of CD14 on human gingival fibroblasts by Arginine-specific cysteine proteinases from Porphyromonas gingivalis leading to down-regulation of lipopolysaccharide-induced interleukin-8 production. Infect Immun. 70(6): 3304-3301. Takahashi, N., Sato, T. and Yamada, T. (2000). Metablolic pathways for cytotoxic end product formation from glutamate- and aspartate-containing peptides by Porphyromonas gingivalis. J Bacteriol. 182(17): 4104-41I0. Takeuchi, T., Kato, N., Watanabe, K. and Morimoto, K. (2000). Mechanism of oxidative DNA damage induction in a strict anaerobe, Prevotella melaninogenica. FEMS Microbiol Lett. I92(I): 133-138. Takeuchi, T., NakayãrY., Kato, N., Watanabe, K. and Morimoto, K. (1999). Induction of oxidative DNA damage in anaerobes. FEBS Lett. 450(3): 178- 180. Tally, F. P., Goldin, B. R., Jacobus, N. V. and Gorbach, S. L. (1911). Superoxide dismutase in anaerobic bacteria of clinical significance. Infect Immun. 16(1): 20-5. Tao, K., Makino, K., Yonei, S., Nakata, A. and Shinagavva, H. (1989). Molecular cloning and nucleotide sequencing of oxyR, the positive regulatory gene of a regulon for an adaptive response to oxidative stress in 235 Escherichia coli: homologies between OxyR protein and a family of bacterial activator proteins. Mol Gen Genet. 218(3):37I-316. Toledano, M.8., Kullik,I., Trinh, F., Baird, P. T., Schneider, T. D. and Storz, G. (1994). Redox-dependent shift of OxyR-DNA contacts along an extended DNA-binding site - a mechanism for differential promoter selection. Cell. 78(5) : 891 -909. Valentine, J. S.,'WertzrD. L., Lyons, T. J., Liou, L,L., Goto, J. J. and Gralla, E. B. (1998). The dark side of dioxygen biochemistry. Cun Opinion Chem Bior.2(2):253-262. van der Hoeven, J. S. and de Jong, M. H. (1984). Continuous culture studies and their relation to the in vivo behaviour of oral bacteria. In: Continuous culture. Biotechnology, medicine and the environment. Ed. Dean, A. C. R., Ellwood, D, C. and Evans, C. G. T. England, Ellis Horwod Limited: 89- 109. Veith, P. D., Gerth, H. T., Slakeski, N. and Reynolds, E. C. (2001). Identification of a novel heterodimeric outer membrane protein of Porphyromonas gingivalis by two-dimensional gel electrophoresis and peptide mass fingerprinting. Eur J Biochem.268: 4748-4151 . Veith, P. D., Talbo, G. H., Slakeski, N., Dashper, S. G., Moore, C., Paolini, R. A. and Reynolds, E. C. (2002). Major outer membrane proteins and proteolytic processing of RgpA and Kgp of Porphyromonas girtgivalis W50. Biochem J. 363(Paft 1): 105-115. Vriesema, A. J. M., Dankert, J. and Zaat, S. A. J. (2000). A shift from oral to blood pH is a stimulus for adaptive gene expression of Streptococcus 236 gordonii CH1 and induces protection against oxidative stress and enhanced bacterial growth by expression of msrA.Infect Immun. 68(3): 1061-1068. Weinberg, 4., Belton, C. M., Park, Y. and Lamont, R. J. (1997). Role of fimbriae in Porphyromonas gingivalis invasion of gingival epithelial cells. Infect Immun. 65(1): 313-316. Weiss, E. I., Shaniztki,8., Dotan, M., Ganeshkumar, N., Kolenbrander, P. E. and Metzger, Z. (2000). Attachmenr. of Fusobacterium nucleatum PKl594 to mammalian cells and its coaggregation with periodontopathogenic bacteria are mediated by the same galactose-binding adhesin, Oral Microbiol Immunol. 15(6): 37 I-317 . Williams, C. H., Jr. (1992). Chemistry and biochemistry of flavoenzymes. Boca Ratón, FL., CRC Press. Wu, J. and Weiss, B. (1991). Two divergently transcribed genes, soxR and soxS, control a superoxide response regulon of Escherichia coli. J Bacteriol. t13(9): 2864-287 t. Xie, H. 4., Cai, S. W. and Lamont, R. J. (1991). Environmental regulation of fimbrial gene expression in Porphyromonas gingivalis. Infect Immun. 65(6): 2265-221L Ximenez-Fyvie, L.4., Haffajee, A. D. and Socransky, S. S. (2000). Comparison of the microbiota of supra- and subgingival plaque in health and periodontitis. J Clin Periodontol. 2l (9): 648-651 . Yesilkaya, H., Kadioglu, P., Gingles, N., Alexander, J. E., Mitchell' T. J. and Andrew, P. W. (2000). Role of manganese-containing superoxide dismutase in oxidative stress and virulence of Streplococcus pneumoniae. Infect Immun. 68(5): 2819-2826. 231 Yi, X., Kot, E. and Bezkorovainy, A. (1998). Properties of NADH oxidase from Lactobacillus delbrueckil ssp bulgaricus. J Sci Food Agric .18: 527 -534. Zhang,J., Dong, H., Kashket, S. and Duncan, M. J. (1999).IL-8 degradation by P o rphy romonas gin g iv alis proteases. Microb Pathogen. 26(5) : 21 5 -280' Zheng,M., Aslund, F. and Storz, G. (1998). Activation of the OxyR transcription factor by reversibte disulfide bond formation. Science. 219(5351): 1718- 1121. Zheng, M., Wang, X., Templeton, L. J., Smulski, D. R., LaRossa, R. A. and Storz, G. (2001). DNA microarray-mediated transcriptional profiling of the Escherichia coli response to hydrogen peroxide. J Bacteriol. 183(15): 4562- 4570. 238