Invest Inve stig igat atio ion n of the the effe effect ct of the the coro corona nall rest restor orat atio ion n quality on the composition of the root canal microflora in teeth with apical periodontitis by means of T-RFLP analysis
G. M. G. Hommez1*, R. Verhelst2*, G. Claeys2, M. Vaneechoutte2 & R. J. G. De Moor1 1
Department of Operative Dentistry and Endodontology, Dental School; and 2Department of Clinical Chemistry, Microbiology and Immunology, Ghent University, Ghent University Hospital, Gent, Belgium
Abstract Hommez GMG, Verhelst R, Claeys G, Vaneechoutte M, De Moor Moor RJG. RJG. Investi Investigat gation ion of the eff effect ect of the corona coronall restor restorati ation on qua qualit lity y on the com compos positi ition on of the roo roott can canal al microflora in teeth with apical periodontitis by means of T-RFLP analysis. International Endodontic Journal , 37 , 819–827, 2004.
Aim To investigate the effect of the radiographic and clinical quality of coronal restorations on the composition of the root canal flora of teeth with necrotic pulps and teet teeth h with with roo roott filling fillingss associ associate ated d with with apical apical periodontitis. Twenty-eight necrotic pulps and 35 Methodology Twenty-eight root filled canals with signs of apical periodontitis were studie studied. d. Bot Both h the coro coronal nal filling filling (pr (presen esence ce of radioradiographically or clinically deficient margins and/or secondary caries) and the root filling (homogeneity and length)) were scored. Bacterial root canal samples were length taken with sterile paper points under rubber dam and usi using ng measur measures es to preven preventt con contam tamina inatio tion. n. A DNA DNA-based nonculture bacterial identification technique was used, use d, nam namely ely termina terminall restri restricti ction on fragme fragment nt length length
polymorphism (T-RFLP) analysis. Results Twelve samples were negative for bacterial DN DNA. A. A tota totall of 33 diff differe erent nt te term rmin inal al restr restric icti tion on fragme fragments nts (TR (TRFs) Fs) wer weree detecte detected. d. The Fusobacterium
nucleatum/Strept Streptococcus ococcus mitis grou group p wa wass th thee mo most st freque frequentl ntly y enco encount untere ered d TRF TRF.. The mea mean n number number of TRFs per necrotic pulp was 6.2 and 5.8 for the groups with acceptable and unacceptable coronal restorations, respectively. This difference was not significant. In the roott filled roo filled gro group, up, these these val values ues (respec (respectiv tively ely,, 5.2 and 8.6) were statistically significantly different (P < 0.05). The following parameters in root filled teeth had no signifi significan cantt infl influen uence ce on the mea mean n num number berss of TRFs TRFs detected: the length and homogeneity of the root filling and the type of tooth (anterior–premolar–molar). Conclusion T-RFLP allowed the rapid assessment of bacterial biodiversity in root canal samples. The technique revealed the presence of bacteria that have rarely been described in the root canals of teeth with apical periodontitis. Biodiversity in the root filled group was high, high, as com compar pared ed with with cul cultur ture-d e-depe epende ndent nt stu studie diess where monoinfections were more frequently reported. Only in root filled teeth did defective coronal restorations have a statis statistical tically ly significan significantt influe influence nce on the mean numbers of detected TRFs per sample.
apical ical perio periodonti dontitis, tis, bacter bacterial ial analy analysis, sis, Keywords: ap endodontics, necrosis, root filled teeth, T-RFLP.
Received 16 April 2004; accepted 17 August 2004
Introduction *First and second author with equal contribution. Corresponde Corres pondence: nce: Geert Homm Hommez, ez, Depart Department ment of Operat Operative ive Dentistry Denti stry and Endodonto Endodontology, logy, Dental Schoo School, l, Ghent University, Ghent Unive University rsity Hospital, Hospital, De Pintelaan Pintelaan 185, B-90 B-9000 00 Gent, Belgium (Tel.: +32 9 240 40 00; fax: +32 9 240 38 51; e-mail:
[email protected]).
[email protected]).
ª 2004 International Endodontic Journal
Coronall leakag Corona leakagee has been recogni recognized zed as a cause cause of failure in root canal treatment (Saunders & Saunders 1994). A number of in in vitro studies have demonstrated the possibility possibility of bacterial leakage along root fillin fillings gs as well as along temporary coronal restorations (Khayat et al. 1993, Deveaux et al. 1999). These findings were
International Endodontic Journal, 37 ,
819–827, 2004
819
T-RFLP analysis and endodontic infections Hommez et al.
supp supple leme ment nted ed by da data ta from from in vivo vivo studie studiess tha thatt indirectly described the effect of coronal leakage (Ray & Tro Trope pe 199 1995, 5, Tro Tronst nstad ad et al. 2000, Homme Hommezz et al. 2002). 200 2). These These studie studiess gav gavee evi eviden dence ce of an increa increased sed incidence of apical periodontitis associated with defective root fillings and with leaking coronal restorations. Bacteria are present in the oral cavity and potentially leak into root canals exposed to the oral environment.
with with api apical cal period periodont ontiti itiss has not yet been invest investiigated. The The ai aim m of th thee pr pres esen entt stud study y wa wass ther theref efor oree to inv invest estiga igate te whe whethe therr it was pos possib sible le usi using ng T-RFLP T-RFLP to ide identi ntify fy the effect effect of the rad radiog iograp raphic hic and cli clinic nical al quality of the coronal restorations on the composition of the root canal microflora in teeth with necrotic pulps and teeth teeth with with roo roott filling fillingss ass associ ociate ated d with with api apical cal
The analysis of these complex bacterial communi communities ties has classically been restricted to conventional culture method met hods. s. The These se met method hodss are time-co time-consu nsumin ming g and laborious. Furthermore, approximately 50% of the oral microflora cannot be cultured (Socransky et al. 1963). Others have shown that the serial dilution anaerobic culture procedure recovers only 20–70% of the microscopic count (Mombelli et al. 1989). Culture-independent mol molecu ecular lar tech techniq niques ues,, freque frequentl ntly y bas based ed on the detection of 16S rRNA genes, have been developed with the aim of includ including ing nonvia nonviable ble,, unc uncult ultiva ivable ble and fastidious fasti dious bacteria. These techn techniques iques inclu include de fluores fluores-situ tu hybridizati cent in si hybridization on (FISH) (FISH) (Langendij (Langendijk k et al. 1995), 199 5), den denatu aturat rating ing gra gradie dient nt gel electro electropho phores resis is (DGGE) (Heuer & Smalla 1997), temperature gradient gel elec electro tropho phoresi resiss (TG (TGGE) GE) (He (Heuer uer & Smalla Smalla 199 1997), 7), tempor tem poral al tem temper peratu ature re gra gradie dient nt gel electro electropho phores resis is (TTGE) (Vasquez et al. 2001) and cloning (Suau et al. 1999, Munson et al. 2002). This study used terminal restriction restri ction fragm fragment ent length length poly polymorph morphism ism (T-RF (T-RFLP) LP) analysis (Liu et al. 1997) to describe the microflora in root root can canal alss wi with th necr necrot otic ic pulp pulps, s, as well well as in root root fillings associated with post-treatment disease. Thus far, only one study used T-RFLP for the characterization of the ora orall bac bacter terial ial microfl microflora ora in the saliva saliva of hea health lthy y subjects subje cts and of patie patients nts with periodontitis periodontitis (Sakamoto et al. 2003). Techniques such as T-RFLP may have the additional addit ional advantag advantagee of high throughp throughput ut and reprodu duci cibi bili lity ty for for mo moni nito tori ring ng ba bact cteri erial al co comm mmun unit itie iess
periodontitis.
(Osborn et al. 2000, Black Blackwood wood et al. 2003), 2003), thus providing a rapid method of finding major differences between communities and testing hypotheses based on a comparison of samples (Blackwood et al. 2003). Cult Cultur ure-b e-bas ased ed info inform rmat atio ion n is avai availa labl blee on the the compos com positi ition on of the bac bacter terial ial mic microfl roflora ora in necr necroti oticc root canals, and in root canal fillings associated with failure failu re (Dahle (Dahle´n & Haapasalo Haapasalo 1998). Molecu Molecular-b lar-based ased techniq tech niques ues wer weree also also use used d to invest investiga igate te the micromicroflora of root canals with apical periodontitis (Conrads et al. 1997, Rolph et al. 2001, Munso Munson n et al. 2002, Siqueira et al. 2002). However, the effect of defective coro corona nall rest restor orat atio ions ns on a poss possib ible le sh shif iftt in th thee co comp mpos osit itio ion n of the the root root ca cana nall micr microfl oflor ora a in tee teeth th
820
International Endodontic Journal, 37 ,
819–827, 2004
Materials and methods Case selection Th Thee 63 pa pati tien ents ts se sele lecte cted d in th this is st stud udy y al alll re recei ceive ved d conventional root canal treatment or retreatment. The pa pati tien ents ts were were he heal alth thy y ad adul ults ts (20 (20 ma male less an and d 43 female fem ales), s), with with an ave averag ragee age of 40.3 40.3 ± 14.0 ye years ars.. This study was approved by the Ethics Committee of the Ghent University Hospital (No. 2003 2003/006) /006).. Infor Informed med consent was obtained from all participants. The samples were taken from previously untreated root canals with necrotic pulps ( n 28) or filled root canalss (retreatment canal (retreatment cases) (n 35). All teeth included in the study had radiographic signs of apical periodontitis visible on a pretreatment radiograph (i.e. at least twice the width of the normal periodontal ligament). Da Data ta on coro corona nall resto restora rati tion onss we were re re reco cord rded ed.. A coronal restoration was categorized as deficient when one of the following criteria was present: radiographic signs of deficient margins on the pretreatment radiograph, clinical signs of a physical defect detected with a probe penetrating the tooth-restoration interface, secondary caries, partial or total loss of restoration. These parameters coincide with the criteria used in a previous study on coronal leakage (Hommez et al. 2002). The following cases were excluded from the study: ¼
¼
Teeth • Teeth
with with a period periodont ontal al pocket pocket ext extend ending ing to the apical third of the tooth. Teeth tha thatt cou could ld not be sui suitab tably ly iso isolat lated ed fro from m the • Teeth gingiva and saliva. • Patients who received antibiotic treatment within the preceding 3 months. • Patients with systemic disease. The quality of an existing root filling was assessed on a pretre pretreatm atment ent rad radiog iograp raph h accord according ing to length length and homogeneity. The length was categorized as acceptable when the root filling terminated 0–2 mm from the root apex. Overextension and underextension of the root filling was scored when the apical extent of the root filling ended, end ed, respect respective ively, ly, bey beyond ond the roo roott apex apex and mor moree
ª 2004 International Endodontic Journal
Hommez et al. T-RFLP analysis and endodontic infections
than 2 mm from the root apex. A root filling was scored homogenous when no voids were present and the root filling was well compacted in the root canal along the whole length of the root filling. Sampling procedure The teet teeth h selected selected for bac bacteri terial al sam sampli pling ng wer weree first first
buffer (Qiagen), samples were incubated for 30 min at 56 C. Next, 200 lL of ethanol was added and DNA was was puri purifie fied d by ad addi ding ng th thee ly lysa sate te to the the Qiag Qiagen en column columnss as descri described bed by the man manufa ufactu cturer. rer. Fin Finall ally, y, the total bacterial DNA was eluted with 100 lL of AE buffer. DNA extracts were stored at ) 20 C. T-RFLP analysis
cleaned clea ned with pum pumice ice and a rub rubber ber dam was placed placed.. Defective coronal fillings and caries, if present, were Defective remove rem oved d and the teet teeth h and rubb rubber er dam were sub subsesequently disinfected with 30% hydrogen peroxide and a 10% iodine tincture, according to the protocol proposed by Mo¨¨ller ller (19 (1966) 66).. After After the tin tinctu cture re had dried, dried, the tooth surface was swabbed with a 5% sodium thiosulphate solution to inactivate the iodine tincture so that remnants of iodine would not influence the bacteriological sample. Subsequently, canal access was gained and enlarged with Gates Glidd Glidden en drill drillss (Dents (Dentsply ply Mail Maillefer, lefer, Ballaigues, Switzerland). If present, the coronal part of the root filling was removed with Gates Glidden drills and the more apical apical part of the root filling filling was par partia tially lly removed using K-Flexofiles (Dentsply Maillefer). Instrumentation menta tion was carried out as close as possi possible ble to 2 mm from the apical constriction. This was checked with an electronic apex locator (Apex Finder AFA; SybronEndo, Orange, CA, USA). A small amount of saline (0.85%) was introduced in the canal with a sterile syringe. The fluid fluid was then mixed with the conten contents ts of the root canal with a root canal file, making pumping movements. The canal fluid was subsequently soaked into a sterile ster ile pap paper er point. point. The pap paper er points points were were direct directly ly transferred transf erred to a tube containin containing g steril sterilee physiologi physiological cal water. wat er. The transp transport ort tub tubes es wer weree transf transferr erred ed to the laboratory within 30 min. The root canal instrumentation tatio n was continued after the sampling procedure.
The forward primer 10f (5¢-AGTTTGATCCTGGCTCAG) an and d the the re reve vers rsee pr prim imer er 534r 34r (5¢-ATTACCGCGGCTGCTGG) CTGCT GG) (Muyz (Muyzer er et al. 1993) 1993),, whi which ch tar target get the 16S rRNA gene (16S rDNA) of the domain Bacteria, were used to amplify part of the 16S rDNA by PCR. A 15-lL PCR mixture mixture con contai tained ned 0.1 lmol L)1 of eac each h primer primer,, 7.5 lL of Pr Prom omeg ega a ma mast ster er mi mix x (P (Pro rome mega ga,, Madiso Mad ison, n, WI, USA USA), ), 1.5 lL of sa samp mple le an and d di dist stil ille led d water. Thermal cycling consisted consisted of an initi initial al denaturation of 5 min at 94 C, followed by three cycles of 1 min at 94 C, 2 min at 50 C and 1 min at 72 C, followed by 35 cycles of 20 s at 94 C, 1 min at 50 C and 1 min 72 C, with a final extension of 10 min at 72 C, and and co cool olin ing g to 10 C. A 20-lL restric restrictio tion n mixture, containing 1 lL PCR product, 1 lL of BstUI (Westburg, (Westb urg, Leiden, the Netherl Netherlands) ands) and 4 lL of the appropriate buffer, was incubated at 60 C for 3 h. Ten microl mic rolitr itree of the restri restricti ction on reacti reaction on was purifie purified d by ethanol precipitation. The obtained pellet was solved in 13.1 lL deion deionized ized formamide (AMRESCO, (AMRESCO, Solon Solon,, OH, USA), 0.1 lL ROX500 and 0.3 lL HD400 GeneScan size size standa standards rds (Appli (Applied ed Biosys Biosystem tems, s, Fos Foster ter Cit City, y, CA, USA) followed by denaturation at 96 C for 2 min and immediate immed iate cooling on ice. The fluore fluorescentl scently y label labelled led termina term inall restrict restriction ion fra fragme gments nts (TR (TRFs) Fs) were were electro electro-phoresed on an ABI PRISM 310 (Applied Biosystems). Figure 1 shows the output of a T-RFLP profile. TRFs with a peak height less than 10% of the highest peak
DNA extraction
were exc were exclud luded ed fro from m the ana analys lysis, is, as clo clonin ning g stu studie diess (unpublished data) indicated that such peaks did not correspond corresp ond with any of the species shown to be present by cloning.
For DNA extraction of bacteria that were collected on pape paperr po poin ints ts,, the the QIAa QIAamp mp DN DNA A mi mini ni kit kit (Qia (Qiage gen, n, Hilden, Germany) was used according to the manufacturer’ss recomm turer’ recommendat endations, ions, with mino minorr modi modificatio fications. ns. After 10-min centrifugation at 5000 g, the pellet was resuspe resu spende nded d in 180 lL of lysis lysis buffer buffer (20 mmo mmoll L)1 Triton). n). Tri Tris–H s–HCl, Cl, pH 8.0 8.0;; 2 mm mmol ol L)1 EDTA; 1.2% Trito )1 Fi Fift fty y un unit itss of mu muta tano noly lysi sin n (2 (25 5 U lL ) (S (Sig igma ma,, Bornem, Belgium) were added and the samples were incubated for 30 min at 37 C. After the addition of 20 lL Prot Protei eina nase se K (20 (20 m mg gm mL L)1) and and 20 200 0 lL AL
ª 2004 International Endodontic Journal
Construction of the T-RFLP library The T-R T-RFLP FLP pat patter tern n obt obtain ained ed fro from m a mix mixed ed sam sample ple consists of the 5¢ terminal BstUI restriction fragments obtained from amplified rDNA of the different species pre present sent.. Theoret Theoretica ically lly the num number ber of pea peaks ks (TRFs) (TRFs) reflects reflects the num number ber of dif differ ferent ent spe species cies present present in a sample. Identification of the peaks in a T-RFLP pattern, in other words assignation of a species name to each
International Endodontic Journal, 37 ,
819–827, 2004
821
T-RFLP analysis and endodontic infections Hommez et al.
30
60
90
120
150
180
210
240
270
300
330
360
390
7 . 5 0 1
420
450
2 . 4 1 4
2000
1500
1000
4 . 9 0 1 7 . 0 2 1
500
5 . 2 8 3
4 . 4 3 1
0 The peaks are numbered according to the number of base pairs in the terminal fragment
Figure 1 Terminal fragment length polymorphism analysis (T-RFLP) of a necrotic root canal with leakage.
TRF, is based on comparison with a library composed of the T-R T-RFLP FLP pattern patternss of wel well-i l-iden dentifi tified ed spec species ies.. Such Such library libra ry T-RFLP patterns consist of a singl singlee TRF, as they are obtain obtained ed from from pur puree cul cultur tures es of a single single specie species. s. Anothe Ano therr mea means ns to obtain obtain T-RFL T-RFLP P pattern patternss of single single speci species es is by ca carry rryin ing g ou outt compu computer ter-a -assi ssist sted ed (i (i.e .e.. virtua virtual) l) restri restricti ction on ana analys lysis is of pub publis lished hed 16S rRN rRNA A sequen sequences ces or by car carryi rying ng out restri restricti ction on ana analys lysis is of cloned 16S rDNA, which is also sequenced to obtain the species name. Data analysis T-RFLP patterns were obtained as table files from the Genescan Analysis software and used in BaseHopper, a software softw are progra program m develo developed ped at our universit university y (Bael (Baelee et al. 2000). Using these sample files containing TRFs (peak values) in base pairs, this program enabled us to construct manually a library which contains one entry for each species and whereby each entry consists of a numeri num ericc value value rep represe resenti nting ng the pea peak k value value in bas basee pairs. pairs. The pea peak k values values in the libra library ry ent entrie riess are the averages averag es of the pea peak k values values obtain obtained ed after after testing testing differ different ent str strain ainss or cloned cloned 16S rRNA genes of each species. Data analysis and statistical analysis were performed in SPSS 11.0.1 (SPSS Inc., Chicago, IL, USA). The mean number of TRFs per sample in the different groups gro ups wer weree analys analysed ed with with the Stu Studen dent’s t’s t-t -test est or anova when more than two groups were present.
Results A total of 12 samples, of which nine in the root filled group, gro up, wer weree PCR PCR-ne -negat gative ive.. Of the 51 PCR PCR-po -posit sitive ive
822
International Endodontic Journal, 37 ,
819–827, 2004
sample sam ples, s, 19 teeth teeth sho showed wed sig signs ns of defecti defective ve corona coronall restorations. Both the pulp necrosis group and the root filled group were divided into subgroups of teeth with accept acc eptabl ablee and una unaccep cceptab table le corona coronall rest restora oratio tions. ns. Tabl Tablee 1 gi give vess an ov over ervi view ew of the the TRFs TRFs an and d the the corresponding species or species groups present in the different patient subgroups. A total of 33 different TRFs were detected. The number of TRFs detected per sample ranged from 0 to 12 for the pulp necrosis group and 0 to 14 for the root filled group. Thirteen TRFs were present in all four groups. Among these, the TRF with a len length gth of 25 bp (br (briefl iefly y TRF TRF25) 25),, design designate ated d as the Atopobium – Treponema Treponema group, group, TRF TRF105 105 (Fusobacterium nucleatum/Streptococcus mitis group), TRF109–110 (an unidentified Veillonella sp.), sp.), TRF TRF111 111 ( Prevotella spp.), TRF120 TRF12 0 (the Cartonella/Peptostreptococcus /Eubacterium group) and TRF410 (the Campylobacter curvus/Dialister sp. group) were present in 19 or more of the patients. All 33 TRFs, except TRF452 ( Prevotella heparinolytica), were present in the root filled group. Only three TRFs: TRF241 TRF24 1 [the Actinomyces Actinomyces israelii israelii/Enter Enterococcus ococcus faecalis faecalis gr grou oup) p),, TR TRF2 F245 45 (Actinomyces sp./Lactobac Lactobacillu illuss casei group/L. crispatus crispatus) and TRF213 TRF213 (un (unide identi ntified fied iso isolat latee ED ED06 06B) B)]] we were re no nott pr pres esen entt in th thee ne necr cros osis is gr grou oup. p. TRF105 TRF10 5 (repre (representin senting g the Fusobacterium Fusobacterium nucle nucleatum atum/ Streptococcus Strept ococcus mitis group group)) was the mos mostt freque frequentl ntly y encountered TRF: (57.7% of the samples), followed by TRF109–110 ( Veillonella sp.) in 28 samples (53.5%). TRF109–110 was the most frequent species (15 of 25 samples) in the pulp necrosis group. In the root filled group, TRF105 (Fusobacterium nucleatum/Streptococcus mitis group) group) was cou counte nted d 17 times (out of 27). The mean number of TRFs (i.e. species) per sample for the ne necr cros osis is gr grou oup p was was 6. 6.21 21 and and 5. 5.82 82 fo forr th thee so soun und d co coro rona nall and and defe defect ctiv ivee coro corona nall resto restora rati tion on gr grou oup, p,
ª 2004 International Endodontic Journal
Hommez et al. T-RFLP analysis and endodontic infections
Table 1 Results Results of the T-RFLP analysis analysis in relation to the tooth subgroups subgroups (n
¼
63)
Groups Necrosis TRF length (bp) Number Negative Species Atopobium parvulum / Cryptobacterium curtum / Olsenella genome genome sp. C1/ Peptostreptococcus lacrimalis / Slackia exigua / Treponema spp. spp. Actinomyces sp./ Lactobacillus spp. sp./ Lactobacillus Capnocytophaga sp. Capnocytophaga gingivalis / C. C. sputigena Capnocytophaga ochracea Fusobacterium nucleatum / Streptococcus mitis group group Veilonella sp. Prevotella sp. Prevotella sp. S. gordonii / Streptococcus Streptococcus milleri group Catonella morbi / Peptostreptococcus magnus /
Filled (retreatment)
No le leak akag age e
Leaka Leakage ge
No le leak akag age e
17 3
11 0
27 6
8 3
63 12
25
7
4
10
3
24
52–53 95 99 102 105
1 1
6
1 1 1 2 14
1
1 1 1 7
3 3 2 3 30
109–110 111 114 115 120
11 6 4 2 8
4 7 4 2 4
5 11
135 200 213 220
2 1
3 1
2 3
10 10 3
Lea Leaka kage ge
3 3 3 2
Total
3
28 19 10 9 26
2 1 1 2
9 6 1 7
P. octavius / Eubacterium Eubacterium saphenum / E. tardum Peptostreptococcus sp. oral clone CK035
Unidentified clone ED52c4 Unidentified isolate ED006B Propionibacterium granulosum / Propionibacterium sp./ unidentified oral bacterium ED52-1 Propionibacterium spp. Propionibacterium acnes / P. P. freudenreichii Peptostreptococcus sp. Staphylococcus sp. sp. oral clone/ Staphylococcus Peptostreptococcus micros Actinomyces israelii / Enterococcus Enterococcus faecalis Actinomyces sp./ Lactobacillus casei group/ L. L. crispatus sp./ Lactobacillus Campylobacter (Bacteroides) gracilis / C. C. showae Lactobacillus catenaformis Pseudomonas stutzeri / Enterobacteriaceae Enterobacteriaceae Citrobacter freundii / unidentified oral bacterium ED25E/ Neisseria elongata /
3
2
222 225 231 238 241 24 245 271 27 383 385–387
1 1 7
2
4
1 3 2
3 1
390
1
2
393 396 399 405 410 414 417 452 A
1 3 1 4 5 3
2 1 1 2 5 5 2 2
2 1 1 1 1 1
4 3 2 15 6 6 4 9 4
9
4
16
1
1 1
3 6 3 13 19 5 3 2 43 346
1
Solobacterium moorei Rothia dentocariosa
Staphylococcus hominis
Selenomonas sputigena Selemonas sp oral clone ED52c21 Campylobacter curvus
Dialister pneumosintes Lactobacillus gasseri Prevotella heparinolytica
Peaks not assigned to any TRF (17 different lengths) Total number of TRFs detected Number of different TRFs detected Mean number of TRFs per sample (negative samples not included)
ª 2004 International Endodontic Journal
87 26 6.21
2 1 2 5 1 2 2 64 23 5.82
1 6 6
1 3 1
1
109 27 5.19
43 24 8.60
International Endodontic Journal, 37 ,
819–827, 2004
823
T-RFLP analysis and endodontic infections Hommez et al.
respectively. This difference was not statistically significant. In the root filling group, the mean number of TRFs was, respectively, 5.19 and 8.60. This difference was statistically significant (P < 0.05). Table 2 presents the number of TRFs in filled root canals according to the quality of the root filling and the tooth type. None of the parameters scored had a significant influence on the number of TRFs detected
The choice of the restriction enzyme used is important. BstUI was chosen, based on in silico silico analysis of 16S rRNA rRN A gen genes es and on the results results of prev previou iouss researc research h (Engeb (En gebret retson son & Moy Moyer er 200 2003), 3), indica indicatin ting g tha thatt thi thiss re rest stri ricti ction on en enzy zyme me wa wass well well suit suited ed fo forr ma maxi xima mall diff differentia erentiation tion between bacter bacterial ial species based on the length of the terminal 5 ¢ restriction fragment of their 16S rDNA, i.e. their TRF. Nevertheless, several species
per sample. Three TRFs (TRF200, 213 and 405) could be linked to a clone or an isolate, but 16S rRNA gene sequencing of both clone and isola isolate te yielded yielded unide unidentified ntified organisms. ism s. Fur Furthe thermo rmore, re, 17 differ different ent TRFs, TRFs, obs observ erved ed in a total number of 43 cases, could not be assigned to any organism. For example, TRF404 was observed in four samples, but no clone or organism could be associated with it.
turned out to have the same TRF. For example, after restriction with Bst UI, the species Atopobium parvulum, Cryptobacteri Crypto bacterium um curtum curtum, Olsenella genomospecies genomospecies C1, Peptostreptococcus lacrimalis, Slackia exigu exiguaa and Treponema sp. were shown to have a TRF with a length of 25 bp. This means that the occurrence of this TRF in a T-RFLP pattern does not allow differentiation between these different species. A tot total al of 33 differen differentt TRF TRFss wer weree detecte detected d in the present study. The number of TRFs per sample ranged from 0 to 12 in necrotic pulps and 0 to 14 in filled root canals can als.. Alt Althou hough gh the ave averag ragee num number ber of TRF TRFss per canal that was detected in filled root canals (5.63) was lower than that in necrotic pulps (6.04), this difference was not stati statistical stically ly significant. significant. These result resultss are in opposi opp ositio tion n to tho those se studie studiess usi using ng cultur culturee met method hodss where it was shown that the microflora in filled root canals exists mostly as monoinfections (Molander et al. 1998, Sundqvist et al. 1998). However, in cases with po poor or qual qualit ity y ro root ot fil filli ling ngs, s, th thee comp compos osit itio ion n of th thee microfl mic roflora ora resembl resembled ed tha thatt of unt untrea reated ted roo roott can canals als.. Th This is di diff ffere erenc ncee is as assu sume med d to be a cons conseq eque uenc ncee of inadequate cleaning (Sundqvist et al. 1998, Cheung & Ho 2001). In this study no evidence of this assumption was seen, as the length and homogeneity of the root filling had no statistically significant influence on the number of TRFs detected in each sample. Enterococci are found in approximately one-third of the filled root canals with apical periodontitis (Mo¨ ller
Discussion Dental periapical diseases are caused primarily by the spread of microorganisms and bacterial elements (toxins ins an and d en endo doto toxi xins ns)) from from the the root root ca cana nall into into the the periapical area. Apical periodontitis is a polymicrobial infection infect ion domin dominated ated by oblig obligately ately anaer anaerobic obic bacter bacteria ia (Dahle´ n & Haapasalo 1998). The number of different species per case has been found to be relatively small, normally between two and eight, and with never more than 20 species in one root canal (Dahle´ n & Haapasalo 1998). This study evaluated the usefulness of a molecular techniq tech nique, ue, nam namely ely T-R T-RFLP FLP for the study study of den dental tal periapical diseases. Table 2 The number number of TRFs in retreated teet teeth h according to
the quality of the root filling and the tooth type ( n
¼
26)
TRFs per sample n
Length of the root filling Acceptable 7 Short 15 Overextended 1 Unknown 3 Homogeneity of the root filling Homogeneous 7 Inhomogeneous 16 Unknown 3 Tooth type Front 6 Premolar 6 Molar 14
Mean
SD
5.9 6.0 7.0 4.7
2.80 3.5 – 0.6
6.6 5.8 4.7
2.5 3.5 0.6
5.3 5.7 6.1
1.4 3.3 3.5
Negative samples were not included in this calculation.
824
International Endodontic Journal, 37 ,
819–827, 2004
1966, Fraser 1974, Molander et al. 1998). Enterococcus faecalis appe appear arss to be extr extrem emel ely y re resi sist stan antt to the the antibacterial agents used during root canal treatment an and d it is on onee of th thee few few mi micro croor orga gani nism smss th that at ca can n surv surviv ivee th thee anti anti-b -bac acte teri rial al effe effect ctss of dr dress essin ing g with with calcium hydroxide (Bystro¨m et al. 1985). Enterococci have also been shown to have the ability to survive in root canals as single organisms without the support of other bacteria (Fabricius et al. 1982). The TRF corresponding to E. faecalis or A. israelii (TRF241) was only pre presen sentt in the roo roott fill filled ed gro group. up. Both bac bacteri teria a wer weree isolated, isola ted, although the former more frequently, frequently, from fill filled ed root root can canals als by other other aut author horss (Mo (Molan lander der et al. 1998,, Sundq 1998 Sundqvist vist et al. 1998, Hancock Hancock et al. 2001).
ª 2004 International Endodontic Journal
Hommez et al. T-RFLP analysis and endodontic infections
On the contrary, Rolph et al. (2001), using molecular method met hods, s, failed failed to det detect ect both. both. The There re is no obv obviou iouss explanation for this difference in incidence of isolation of E. E. faecalis between the studies. No information was available regarding the (possibly leaking) temporary fillings or the number of visits spent on the original root canal, which can also influence the presence of E. faecalis faecalis (Siren et al. 1997 1997). ). The presen presence ce of
microorga microo rganis nisms ms were pres present ent but tha thatt the sam sampli pling ng procedure did not pick-up the bacteria in some parts of th thee root root ca cana nal. l. It sh shou ould ld be note noted d that that mo more re nega negati tive ve sa samp mple less we were re fo foun und d with with cult cultur uree-ba based sed methods metho ds (Mola (Molander nder et al. 1998: 1998: 32. 32.0%, 0%, Che Cheung ung & Ho 2001: 33.3%, Sundqvist et al. 1998: 55.6%) than with with the DNA tech techniq nique ue use used d in the presen presentt stu study dy (19.0% (19 .0%)) and in oth other er studie studiess (Co (Conra nrads ds et al. 1997:
defective coronal restorations did not seem to have effect on the presence of TRF241. In teeth with necrotic pulps, TRF 241 was absent. The incidence of TRF 241 was about the same for the samples taken from root filled teeth with unacceptable and acceptable coronal restorations (18.5 and 12.5% of the samples, respectively). No stud study y has has ye yett inve invest stig igat ated ed the the ef effe fect ct of the the radiographic and clinical quality of the coronal restorations on the composition of the root canal microflora in teeth teeth with with necro necroti ticc pu pulp lpss an and d tee teeth th wi with th po post st-treatment disease in vivo. No specifi specificc species or group of speci species es wa wass foun found d that that indi indica cated ted le leak akag age. e. It was was interesting to find that in root filled teeth with defective coronal coro nal rest restora oratio tions, ns, the mean mean number number of TRF TRFss per sample detected was significantly higher than in other cases. Although care should be taken when interpreting this result because of different sample sizes, this difference in the mean number of TRFs per sample from teeth with unacceptable and acceptable coronal restorations could be explained by ingress of bacteria that increased the diversity of the microflora. Nevertheless, no similar significant difference could be detected in the samples sampl es from necrotic pulps. Care should be taken when defining defective restorations (restorations that were scored radiographically and/or clinically unacceptable in this study) as restoration rationss with with cor corona onall leakag leakage. e. A minute minute gap at the coronal or radicular-restoration interface, which would not be visible on the radiograph, could be sufficient to
17.6%,, Sique 17.6% Siqueira ira et al. 2002: 2002: 0%) 0%),, alt althou hough gh Rolph Rolph et al. (2001), who also used T-RFLP as well, reported 31.7% of the root canals to be negative. This study is the first to use T-RFLP to investigate the compositio compo sition n of the root canal mic microflor roflora. a. T-RFLP T-RFLP is most commonly used for systematic community analysis of environmental samples (Marsh 1999, Kitts 2001). The ad adva vant ntag agee of TT-RF RFLP LP is a high higher er th thro roug ughp hput ut and and reproducibi reprod ucibility lity in monit monitoring oring bacterial bacterial communities communities (Osborn et al. 2000 2000,, Black Blackwood wood et al. 2003) 2003).. In addition, addition, T-RFLP was found to be very reproducible. A drawback of thi thiss tech techniq nique ue howeve however, r, is the som someti etimes mes lim limite ited d discriminatory power, as different species can present with the same TRF. Nevertheless, this approach allowed for the detection of bacterial species never previously found in endodontic infections by culture and was in agreem agr eement ent with with the find finding ingss of other other report reportss usi using ng molecular molecu lar techni techniques ques (Conr (Conrads ads et al. 1997, Sique Siqueira ira et al. 2000, Jung et al. 2001, Rolph et al. 2001).
permit bacterial penetration. Severa Sev erall exp explan lanati ations ons are pos possib sible le for the fact fact tha thatt 12 sa samp mple less prod produc uced ed ne nega gati tive ve re resu sult ltss with with th thee T-RFLP technique. As the majority of these negative samples were found in the root filled group, it is not inconc inconceiv eivabl ablee tha thatt som somee of the cas cases es wer weree hea healin ling g and and th that at a rest restri rict cted ed am amou ount nt or no ba bact cteri eria a were were present pres ent in the root root can canal. al. Api Apical cal rad radiol ioluce ucenci ncies es are not not neces necessa sari rily ly ca caus used ed by intr intrac acan anal al infe infect ctio ion. n. A radiolucency may refer to a periapical cyst (Nair et al. 1996 1996), ), a forei foreign gn bo body dy reac reacti tion on (Nai (Nairr et al. 1990, Sj Sjo o¨gre gren n et al. 1995) or scar tissue (Bhaskar 1966). It may also have been possible that errors in sampling resulted in negative samples. It was also possible that
filled canals associated with post-treatment disease, a high higher er dive diversi rsity ty in th these ese sa samp mple less was was fo foun und. d. No statistical differences with regard to biodiversity could be assess assessed ed between between root can canals als with with necr necroti oticc pul pulps ps and filled root canals. In root filled teeth, a significantly higher mean number of TRFs per sample was detected in samples with defective coronal restorations.
ª 2004 International Endodontic Journal
Conclusion T-RFLP, a DNA-based nonculture-dependent approach madee possib mad possible le the rap rapid id ass assessm essment ent of the bac bacteri terial al biodiversity of necrotic pulps and filled root canals with ap apic ical al peri period odon onti titi tiss an and d re reve veal aled ed th thee pr prese esenc ncee of bacteria that have rarely been described in connection with this pathosis. Compared with other, culture-based studies, which usually indicate monoinfection of root
Acknowledgements This study was partly funded by the European Society of Endodontology (ESE) research Grant 2001 and the Fonds voor Wetenschappelijk Onderzoek (FWO), grant no. 1.5.126.04.
International Endodontic Journal, 37 ,
819–827, 2004
825
T-RFLP analysis and endodontic infections Hommez et al.
References Baele M, Baele P, Vaneechoutte M et al. (2000) Application of tDNA-PCR tDNAPCR for the iden identifica tification tion of ente enterococc rococci. i. Journal Journal of Clinical Microbiology 38 , 4201–7. Bhaskar SN (1966) Periapical lesions – types, incidence and clinical features. Oral Surgery Oral Medicine Oral Pathology 21, 657–71. Blackwood CB, Marsh T, Kim S-H, Paul EA (2003) Terminal restriction fragment length polymorphism data analysis for quantitative comparison of microbial communities. Applied and Environmental Microbiology 69 , 926–32. Bystro¨¨m m A, Claesson R, Sundqvist G (1985) The antibacterial effect of camphorated paramonochlorphenol, camphorated phenol phen ol and calcium hydroxide hydroxide in the treatment treatment of infec infected ted root canals. Endodontics and Dental Traumatology 1 , 170–5. Cheung GSP, Ho MWM (2001) Microbial flora of root canaltreated teeth associated with asymptomatic periapical radiolucent oluce nt lesions. lesions. OralMicrobio OralMicrobiologyand logyand Immunology Immunology 16, 332– 332–7. 7. Conrads G, Gharbia SE, Gulabivala K, Lampert F, Shah HN (199 (1997) 7) Th Thee use use of a 16S 16S rDN rDNA A dire direct cted ed PCR PCR for for the the detection of endodontopathogenic bacteria. Journal of Endodontics 23 , 433–8. Dahle´ n G, Haap Haapas asal alo o M (199 (1998) 8) Micro Microbi biol olog ogy y of apic apical al periodontitis. periodont itis. In: Ørstavik D, Pitt Ford TR, eds. Essential Endodontology. Prevention and Treatment of Apical Periodontitis, 1st edn. Oxford, UK: Blackwell Science Publications, pp. 106–30. Deveaux E, Hildelbert P, Neut C, Romond C (1999) Bacterial leakage of Cavit, IRM, TERM, and Fermit: a 21-day in vitro study. Journal of Endodontics 25 , 653–9. Engebr Eng ebrets etson on JJ, Moyer Moyer CL (20 (2003 03)) Fideli Fidelity ty of restri restricti ction on endonuclea endon ucleases ses in deter determini mining ng micro microbial bial diver diversity sity by terminal-rest mina l-restricti riction on fragm fragment ent leng length th polym polymorphi orphism. sm. Applied Environmental Microbiology 69 , 4823–9. ¨ hman Fabr Fabric iciu iuss L, Da´hlen len G, O hman AE, Mo¨¨ller ller AJR (19 (1982) 82) Predominant indigenous oral bacteria isolated from infected root canals after varied times of closure. Scandinavian Journal of Dental Research 90 , 134–44. Fraser JG (1974) Chelating agents: their softening effect on root canal dentin. Oral Surgery Oral Medicine Oral Pathology 37, 803–11. Hancock HH, Sigurdsson A, Trope M, Moiseiwitsch J (2001) Bacteria isolated after unsuccessful endodontic treatment in a North American population. population. Oral Surge Surgery ry Oral Medicine Oral Pathology Oral Radiology and Endodontics 91, 579–86. Heuer Heu er H, Smalla Smalla K (19 (1997) 97) App Applic licati ation on of denatu denaturat rating ing gradient gel electrophoresis (DGGE) and temperature gradient gel electropho electrophoresis resis (TGGE) for study studying ing soil micro microbial bial communities. In: Van Elsas JD, Wellington EMH, Trevors JT, eds. Modern Soil Microbiology. New York: Marcel Dekker, pp. 353–73. Hommez GMG, Coppens CRM, De Moor RJG (2002) Periapical health related to the quality of coronal restorations and root fillings. International Endodontic Journal 35 , 680–9.
826
International Endodontic Journal, 37 ,
819–827, 2004
Jung I-Y, Choi B-K, Kum K-Y et al. (2001) Identification of oral spirochetes spirochetes at the species level and their associatio association n with other bacteria in endodontic infections. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontics 92, 329–34. Khayat Kha yat A, Lee S-J, S-J, Tor Torabi abine nejad jad M (1993 (1993)) Human Human saliva saliva penetrati penet ration on of coronally coronally unsealed unsealed obturated obturated root canal canals. s. Journal of Endodontics 19 , 458–61. Kitts CL (2001) Terminal restriction fragment patterns, a tool for comparing microbial communities and assessing community dynamics. Current Issues in Intestinal Microbiology 2 , 17–25. Langendijk PS, Schut F, Jansen GJ et al. (1995) Quantitative fluorescenc fluore scencee in situ hybri hybridiza dization tion of Bifidobacterium spp. wit with h genusgenus-spe specifi cificc 16S 16S rRN rRNA-t A-targ argete eted d probes probes and its applicati appli cation on in faecal samples. Applied Applied and Envir Environme onmental ntal Microbiology 61 , 3069–75. Liu W, Marsh TL, Cheng H, Forney LJ (1997) Characterization of microbial microbial diversity diversity by deter determini mining ng terminal terminal restri restriction ction fra fragme gment nt len length gth polymo polymorph rphism ism of genes genes encodi encoding ng 16S 16S rRNA. Applied and Environmental Microbiology 63 , 4516–22. Marsh TL (1999) Terminal restriction fragment (T-RFLP), an emerging method for characterizing diversity among homologous populations of amplification products. Current Opinion in Microbiology 2 , 323–7. Molander A, Reit C, Dahle´ n G, Kvist T (1998) Microbiological status of root-filled teeth with apical periodontitis. International Endodontic Journal 31, 1–7. Mo¨¨ller ller AJR (1966) Microbiological examination of root canals and periap periapica icall tis tissue suess of human human teeth. teeth. Met Method hodolo ologic gical al studies. Odontologisk Tidskrift 74 , 1–380. Momb Mombel elli li A, Mind Minder er CE, CE, Gusb Gusber erti ti FA, FA, Lang Lang NP (198 (1989) 9) Reproducibility of microscopic and cultural data in repeated subgingival plaque samples. Journal of Clinical Periodontology 16, 434–42. Munson MA, Pitt-Ford T, Chong B, Weightman A, Wade WG (2002) (200 2) Molec Molecular ular and cultural cultural analysis of the micro microflora flora associated associ ated with endodonti endodonticc infec infection tions. s. Journal Journal of Denta Dentall Research 81 , 761–6. Muyzer G, de Waal EC, Uitterlinden AG (1993) Profiling of complex microbial populations by denaturing gradient gel ele electr ctroph ophore oresis sis analys analysis is of polyme polymeras rasee cha chain in reacti reactiononamplified genes coding for 16S rRNA. Applied and Environmental Microbiology 59, 695–700. Nair PNR, Sjo¨¨gren gren U, Krey G, Kahn Kahnberg berg K-E, Sund Sundqvist qvist G (1990) (199 0) Intr Intraradi aradicular cular bacteria and fungi fungi in root-filled root-filled,, asymptomatic human teeth with therapy-resistant periapica icall les lesion ions: s: a lon long-t g-term erm lig light ht and ele electr ctron on mic micros roscop copic ic follow-up study. Journal of Endodontics 16 , 580–8. NairPNR, PajarolaG, PajarolaG, Schroeder Schroeder HE (199 (1996) 6) Typesand incidence incidence of human periapical lesions obtained with extracted teeth. Oral Surgery Oral Medicine Oral Pathology 81, 93–102. Osborn AM, Moore ER, Timmis KN (2000) An evaluation of termin terminalal-rest restric rictio tion n fra fragme gment nt length len gth polymo polymorph rphism ism (T-RFLP) (T-R FLP) analysis for the study of microbial microbial communi community ty
ª 2004 International Endodontic Journal
Hommez et al. T-RFLP analysis and endodontic infections
structure and dynamics. structure dynamics. Environmental Environmental Microbiology 2, 39–50. Ray HA, Trope M (1995) Periapical status of endodontically treated teeth in relation to the technical quality of the root filling and the coronal restoration. International Endodontic Journal 28 , 12–8. Rolph Rolph HJ, Lennon Lennon A, Rig Riggio gio MP et al. (20 (2001) 01) Molecu Molecular lar identification of microorganisms from endodontic infections. Journal of Clinical Microbiology 39 , 3282–9.
Sjo¨gren U, Sundqvist G, Nair PNR (1995) Tissue reaction to gut guttata-per percha cha partic particles les of variou variouss sizes sizes when when implan implanted ted European Journal of Oral subcutane subcu taneously ously in guinea guinea pigs. European Sciences 103, 313–21. Socransky SS, Gibbons RJ, Dale AC, Bortnick L, Rosenthal E, MacDon Mac Donald ald JB (19 (1963 63)) The The microb microbiot iota a of the gin gingiv gival al crevice in man. 1. Total microscopic and viable counts and counts coun ts of speci specific fic organ organisms. isms. Archive Archivess of Oral Biology 8, 275–80.
Sakamo Sakamoto to M, Takeuc Takeuchi hi Y, Umeda Umeda M, Ish Ishika ikawa wa I, Benno Benno Y (2003) Application of terminal RFLP-analysis to characterize oral oral bac bacter terial ial flora in saliva saliva of hea health lthy y sub subjec jects ts and patients with periodontitis. Journal of Medical Medical Microbiolo Microbiology gy 52, 79–89. Saunde Sau nders rs WP, WP, Sau Saunde nders rs EM (19 (1994 94)) Cor Corona onall leakag leakagee as a cause of failure in root-canal therapy: a review. Endodontics and Dental Traumatology 10 , 105–8. Siq Siquei ueira ra JF, Rocas Rocas IN, Oli Olivei veira ra JCM JCM,, San Santos tos KRN (2 (2000 000)) Detection of Treponema Treponema denticola in endodontic infections by 16S rRNA gene directed polymerase chain reaction. Oral Microbiology and Immunology 15, 335–7. Siqueira JF, Roˆc¸¸as as IN, Moraes SR, Santos KRN (2002) Direct amplification of rRNA gene sequences for identification of selected oral pathogens in root canal infection. International Endodontic Journal 35 , 345–51. Sir Siren en EK, Haapasal Haapasalo o MPP, MPP, Ranta Ranta K, Salmi Salmi P, Ker Kerosu osuo o ENJ (1997) (199 7) Micro Microbiol biologica ogicall findings findings and clini clinical cal treat treatment ment procedures in endodontic cases selected for microbiological investigation. International Endodontic Journal 30 , 91–5.
Suau A, Bon Suau Bonnet net R, Sut Sutren ren M et al. al. (19 (1999 99)) Dir Direct ect ana analys lysis is of genes encodin encoding g 16S rRNA from complex communities communities revealss many novel molecular species within the human reveal human gut. Applied Applied and Environmen Environmental tal Micr Microbiolo obiology gy 65, 4799 4799– – 807. Sundqvist G, Figdor D, Persson S, Sjo¨gren U (1998) Microbiologic analysis of teeth with failed endodontic treatment and the outcome of conservative re-treatment. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontics 85, 86–93. Tronstad L, Asbjørnsen K, Døving L, Pedersen I, Eriksen HM (2000) Influence of coronal restorations on the periapical health of endodontically treated teeth. Endodontics and Dental Traumatology 16 , 218–21. Vasquez AS, Ahrne B, Petterson B, Molin G (2001) Temporal temperature gradient gel electrophoresis (TTGE) as a tool for identificat ident ification ion of Lactobacillus Lactobacillus casei, Lactobacillus Lactobacillus paracasei, Lactobacillus zea and Lactobacillus rhamnosu. Letters of Applied Microbiology 32, 215–9.
ª 2004 International Endodontic Journal
International Endodontic Journal, 37 ,
819–827, 2004
827
