The bacteriology of chronic venous leg ulcer examined by culture-independent molecular methods

Wound Repair and RegenerationVolume 18, Issue 1 p. 38-49 Free Access The bacteriology of chronic venous leg ulcer examined by culture-independent molecular methods Trine R. Thomsen PhD, Trine R. Thomsen PhD Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, Denmark The Danish Technological Institute, Chemistry and Water Technology, Århus, DenmarkSearch for more papers by this authorMartin S. Aasholm MSc, Martin S. Aasholm MSc Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, DenmarkSearch for more papers by this authorVibeke B. Rudkjøbing BSc, Vibeke B. Rudkjøbing BSc Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, DenmarkSearch for more papers by this authorAaron M. Saunders PhD, Aaron M. Saunders PhD Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, Denmark The Danish Technological Institute, Chemistry and Water Technology, Århus, DenmarkSearch for more papers by this authorThomas Bjarnsholt PhD, Thomas Bjarnsholt PhD Department of International Health, Immunology and Microbiology, University of Copenhagen, Copenhagen, DenmarkSearch for more papers by this authorMichael Givskov Dr. Tech, Michael Givskov Dr. Tech Department of International Health, Immunology and Microbiology, University of Copenhagen, Copenhagen, DenmarkSearch for more papers by this authorKlaus Kirketerp-Møller MD, Klaus Kirketerp-Møller MD Copenhagen Wound Healing Center, Bispebjerg Hospital, Bispebjerg, DenmarkSearch for more papers by this authorPer H. Nielsen PhD, Per H. Nielsen PhD Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, DenmarkSearch for more papers by this author Trine R. Thomsen PhD, Trine R. Thomsen PhD Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, Denmark The Danish Technological Institute, Chemistry and Water Technology, Århus, DenmarkSearch for more papers by this authorMartin S. Aasholm MSc, Martin S. Aasholm MSc Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, DenmarkSearch for more papers by this authorVibeke B. Rudkjøbing BSc, Vibeke B. Rudkjøbing BSc Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, DenmarkSearch for more papers by this authorAaron M. Saunders PhD, Aaron M. Saunders PhD Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, Denmark The Danish Technological Institute, Chemistry and Water Technology, Århus, DenmarkSearch for more papers by this authorThomas Bjarnsholt PhD, Thomas Bjarnsholt PhD Department of International Health, Immunology and Microbiology, University of Copenhagen, Copenhagen, DenmarkSearch for more papers by this authorMichael Givskov Dr. Tech, Michael Givskov Dr. Tech Department of International Health, Immunology and Microbiology, University of Copenhagen, Copenhagen, DenmarkSearch for more papers by this authorKlaus Kirketerp-Møller MD, Klaus Kirketerp-Møller MD Copenhagen Wound Healing Center, Bispebjerg Hospital, Bispebjerg, DenmarkSearch for more papers by this authorPer H. Nielsen PhD, Per H. Nielsen PhD Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Aalborg, DenmarkSearch for more papers by this author First published: 07 January 2010 https://doi.org/10.1111/j.1524-475X.2009.00561.xCitations: 104 Reprint requests: Trine R. Thomsen, Department of Biotechnology, Chemistry and Environmental Engineering, Aalborg University, Sohngaardsholmsvej 49, DK-9000 Aalborg, Denmark. Tel: 45 7220 1828;Fax: 45 9814 1808;Email: [email protected] AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract The bacterial microbiota plays an important role in the prolonged healing of chronic venous leg ulcers. The present study compared the bacterial diversity within ulcer material from 14 skin graft operations of chronic venous leg ulcers using culture-based methods and molecular biological methods, such as 16S rRNA gene sequencing, fingerprinting, quantitative polymerase chain reaction, and fluorescence in situ hybridization. Each wound contained an average of 5.4 species but the actual species varied between wounds. The diversity determined by culture-based methods and the molecular biological methods was different. All the wounds contained Staphylococcus aureus, whereas Pseudomonas aeruginosa was in six out of 14 wounds. Molecular methods detected anaerobic pathogens in four ulcers that were not detected with anaerobic culture methods. Quantitative polymerase chain reaction was used to compare the abundance of S. aureus and P. aeruginosa at different locations in the ulcers and their numbers varied greatly between samples taken at different locations in the same ulcer. This should be considered when ulcers are investigated in routine clinical care. The differences between the results obtained with culture-based and molecular-based approaches demonstrate that the use of one approach alone is not able to identify all of the bacteria present in the wounds. Chronic venous leg ulcers (CVLU) are a debilitating and often painful disease that affects approximately 1% of the world's population.1,2 Apart from the human consequences, the treatment of wounds is expensive; in Denmark alone, wound treatment has been estimated to cost approximately two billion Danish kroner per year (∼US$360 million),2 and in the United Kingdom, France, and Germany an estimated 1.5–2% of the annual healthcare budget.3,4 The conditions leading to a CVLU are not fully understood; however, the primary cause is most likely insufficient valvular function of the veins in the legs causing increased hydrostatic pressure leading to edema of the subcutaneous tissue, which predispose to ulceration. This is linked to old age, obesity, height of the person, hereditary increased risk, number of births (more births lead to increased risk) and occupations in which the person is mainly standing. By removing edema with compression therapy, most CVLU will heal, but a number of ulcers will not despite effective treatment. In these cases, a well-documented and effective treatment is surgical debridement and split skin transplant.2 Other treatments like topical negative pressure therapy have been found useful. Maggot debridement therapy have also proved promising, which involves having larvae from the fly Lucilia sericata removing necrotic tissue and bacteria from the wound, and in this way aiding the wound healing process.5 One of the factors affecting the effectiveness of wound healing therapies is the specific microorganisms that colonize the CVLU.6 For example, the presence of Pseudomonas aeruginosa can retard the healing of wounds due to their ability to form biofilms.6 Many studies describe biofilm as an important factor for the chronic behavior of chronic wounds,6–10 and the spatial organization of these biofilms in a wound might be complex due to, for example, variations in environmental conditions and population composition.11 Initial experiments by Bjarnsholt et al.6 showed that P. aeruginosa in CVLU were assembled in microcolony-based structures unevenly distributed across the wound surface, and this uneven distribution might lead to insufficient sampling and wrong diagnosis.6 Until recently, the bacteria associated with CVLU have only been examined by culture-dependent methods by taking a swab or biopsy from the wound and using it as inoculate for various bacterial cultures. The emergence of molecular biology methods has illustrated that culture-dependent methods often underestimate the bacteria present, and especially ulcers with slow growing, fastidious, or anaerobic microbes.9,12–14 Davies et al.15 found that 40% of the organisms identified from CVLU by molecular biological methods could not be identified by culture-dependent methods, although most were species that are normally considered culturable. The purpose of this study was to investigate the microbial diversity of chronic ulcers with molecular biological methods and to compare these results with the conventional culture-dependent techniques. Furthermore, the spatial organization of bacteria in CVLU was examined. MATERIALS AND METHODS Patient population, sampling, and DNA extraction The excision of biopsies and swabs of the wounds for culture-dependent and -independent experiments was performed by Copenhagen Wound Healing Center, Bispebjerg Hospital (Copenhagen, Denmark). Samples were obtained from patients diagnosed with chronic venous leg wounds just before surgical debridement and split skin transplant. In total, chronic wounds from 14 patients were investigated (named as wound A–N). The patients' age, sex, antibiotic treatment, dressings at the time of sampling, and additional information are described in Table 1. Patients with wound B, F, H, and K were also diagnosed with diabetes mellitus. Table 1. Summary of patient data* Wound Age ofpatient Sex Treatment of samplebefore extraction Antibiotic treatment Dressing at timeof sampling Durationof ulcer Additionalinformation A 85 Male DNA extracted from the entire wound None Nonsilver 12 months B 76 Male DNA extracted from the entire wound None Nonsilver 6 months Diabetic C 54 Male DNA extracted from the entire wound None Aquacell Ag Years D 87 Female DNA extracted from the entire wound None Nonsilver 4 months E 85 Female Wound was cut into five parts and DNA extracted separately None Biatain AG 7 months F 71 Female Wound was cut into five parts and DNA extracted separately Sulfametizole due to urinary tract infect Biatain AG 5 months Diabetic G 88 Female DNA was extracted from six biopsies across the wound None Biatain AG 4 years H 82 Male DNA was extracted from six biopsies across the wound None Nonsilver 6 months Diabetic I 81 Female DNA was extracted from six biopsies across the wound Phenoxymethylpenicillin until 2 months before sampling Nonsilver 4 years J 78 Female DNA was extracted from six biopsies across the wound Phenoxymethyl-penicillin Nonsilver 6 months K 65 Male DNA was extracted from four biopsies across the wound None Nonsilver 6 months Diabetic, impetigo L 85 Female DNA was extracted from four biopsies across the wound None Biatain AG 7 months M 69 Female DNA was extracted from four biopsies across the wound None Nonsilver 6 months N 46 Male DNA was extracted from four biopsies across the wound None Nonsilver 3 years Sample from Achilles tendon Average age 75.2 * All DNA extractions were done using a DNeasy Blood and tissue kit except for the samples from wound F and wound E (center), which was extracted with an E.Z.N.A. Tissue DNA kit due to their greater size. Registered antibiotic treatment 3 months before sampling is mentioned. All ulcers were chronic and nonhealing despite optimal wound care and compression therapy. The duration of the ulcers are shown in Table 1. The patients were not receiving antibiotic treatment during the three months before sampling with three exceptions: Patient F was receiving sulfonametizole at the time of sampling, and patients J and I had received phenoxymethylpenicillin up until 3 days and 2 months before sampling, respectively. Five of the patients' wounds had been dressed with a silver-releasing dressing in the period before sampling (patient C, E, F, G, and L). The samples were collected with the patients' acceptance and in accordance with the biomedical project protocol (KA-20051011) approved by the Danish scientific ethical board. On the day of surgery, the area surrounding the ulcer was swabbed with chlorhexidine in 70% alcohol but the surface of the ulcer was not disturbed. The excised wound material from the patient was transferred to a sterile Greiner tube and stored at −20 °C until DNA extraction. Before DNA extraction, the frozen wounds were thawed and cut to smaller pieces using sterile disposable scalpels. The total DNA content of wound F and E was extracted using an E.Z.N.A Tissue DNA kit (Omega Bio-Tek, VWR, Herlev, Denmark). Other wounds were cut to smaller pieces and were extracted using a DNeasy Blood and Tissue kit from Qiagen (Hilden, Germany). Both kits are based on proteinase K digestion for 2–4 hours. Culture analysis Identification of bacteria from the wounds by culturing was performed by the Department of Clinical Microbiology, Hvidovre Hospital, according to their standard protocols. Tissue samples were transported in sterile containers and swabs were transported in Stuart medium. Anaerobe culturing was performed on anaerobe plates (Statens Serum Institute [SSI], Copenhagen, Denmark) in a CO2 atmosphere at 37 °C for 2 and 4 days. Aerobe culturing was performed on horse-blood agar (SSI) and Blue plates (SSI) for 1 and 2 days, respectively. 16S rRNA gene amplification The 16S rRNA genes were amplified by polymerase chain reaction (PCR) using Taq DNA polymerase with primers targeting conserved domains. The primers were 8F16 and 1390R17 and the samples were amplified according to Thomsen et al.18 Negative controls including water and PCR mix were included for every five samples and were always negative indicating that there was no contamination of the reagents. Stringent procedures were generally used to avoid contamination, e.g., by using a PCR cabinet with UV light and all DNA handling was carried out with aerosol filter pipette tips to avoid cross contamination. Cloning, sequencing, and phylogenetic studies The amplified 16S rRNA gene products were purified with a Qiaquick PCR purification kit (Qiagen), according to the manufacturer's instructions. Cloning was performed using a TOPO TA Cloning® kit (Invitrogen, Taastrup, Denmark) for sequencing. Plasmids were purified using a Fastplasmid mini kit (Eppendorf, Horsholm, Denmark) and purified plasmids were amplified using M13 primers to test for inserts with the correct length. The plasmids were sequenced by Macrogen Inc. (Seoul, South Korea) using the M13F primer. The closest relative of the clones were identified by performing a BLAST search of the sequences at http://www.ncbi.nlm.nih.gov/blast. At least one representative clone from each species was additionally sequenced using the M13R primer, in order to obtain consensus sequences covering the entire length of the DNA fragments. Checks for chimeric sequences were conducted using the program BELLEROPHON.19 The ARB software20 was used for the alignment of imported sequences with the FastAligner tool, and alignments were subsequently refined manually and phylogenetic analysis was performed. Only unambiguously aligned sequences were used for the calculation of trees using distance matrix, parsimony, and maximum likelihood approaches using default settings in the ARB software. The Bacteria sequence conservation filter of the ssu_jan04_corr_opt ARB database [available at http://www.arb-home.de]) in addition was applied. Phylogenetic trees were initially constructed using the consensus sequences representing the different groups of bacteria. Subsequently, partial sequences were added to the existing consensus trees by the "add species to existing tree" function in the ARB software. Priorly, a filter was carried out to define which positions to be used in adding the partial sequences (data not shown). Generally, the results obtained by the NCBI Blast Search corresponded well to the phylogenetic identifications. The coverage ratio (C) for each of the clone libraries were calculated with where n1 is the number of operational taxonomic units (OTUs) containing only one sequence and N is the total number of clones analyzed.21 Denaturant gradient gel electrophoresis (DGGE) fingerprinting Amplification of samples for DGGE was performed using primers 341F-GC 22 and 907R.17 The PCR products were run on 8% polyacrylamide gels containing denaturant gradients of 30–70%, in 1 × TAE buffer at 100 V overnight using the D-GENE™ gel system (Biorad) and stained with SYBR Gold (Invitrogen). The most intensive DGGE bands were excised and prepared for sequencing. The excised bands were reamplified with PCR, and the PCR products were thereafter purified using a NucleoSpin Extract II Machery Nagel and sequenced commercially by Macrogen Inc. Quantitative PCR (qPCR) Pure culture DNA was extracted using a FastDNA® Spin Kit for Soil (MP Biomedicals, LLC, Illkirch, France), according to the manufacturer's instructions. qPCR targeting the nuc gene23 and oprL gene24 was used to measure the amount of Staphylococcus aureus and P. aeruginosa, respectively. For each determination, triplicates of 20 μL reactions were run with each containing: 12.5 μL Brilliant® II SYBR® Green qPCR Master mix (Stratagene, AH Diagnostics, Aarhus, Denmark), 25 μg BSA, 10 μM of each primer, and 0.75 μM reference dye and 5 μL of template or standard. Reactions were run on an Mx3005P (Stratagene) for 10 minutes at 95 °C, 40 cycles of 30 seconds at 95 °C, 30 seconds at 62 °C (nuc)/ 62 °C (oprL), 60 seconds at 72 °C and 15 seconds, and SYBR data capture at 80 °C (nuc)/ 82 °C (oprL). For S. aureus, the specific product was separated at 79 °C and for P. aeruginosa at 90 °C. The specificity of the PCR reactions performed for each run was confirmed by the melting curve analysis and gel electrophoresis. Standard curves were prepared from serial dilutions of S. aureus (DSM 6148) and P. aeruginosa (DSM 1253) genomic DNA (5 × 106–5 × 101) in AE buffer (Qiagen). The limit of detection was 100 gene copies per PCR. Fluorescence in situ hybridization (FISH) After removal from the patient, the tissue sample was transferred to 4% neutral formaldehyde buffer and embedded in paraffin wax, cut into 4-μm–thick slides, and stored at room temperature. Before the hybridization, the paraffin was removed by xylene. The slides were treated using a Histology FISH Accessory Kit from DAKO cytomation according to the protocol. Hybridization was performed by covering the slide with 20 μL of hybridization buffer containing 0.9 M NaCl, 0.02 M Tris/HCl, 0.01% SDS, and formamide, depending on the requirement of the probes and probe mix (5 ng/μL). The probes used were an EUB mix (EUB-338,25 EUB II-338,26 and EUB III-33826) targeting most Bacteria; BET42a with GAM42a competitor27 targeting most Betaproteobacteria; a mix of LGC354b, LGC354A, and LGC354C28 targeting the Firmicutes, and probe Sau29 targeting S. aureus. For more information about the probes, consult probeBase.30 Lastly, the slides where treated with Vectashield hardset mounting medium with DAPI (4′,6-diamidino-2-phenylindole). Unspecific binding was examined by applying Non-EUB probes on a slide as described above. This revealed sporadic nonspecific binding but only with little signal intensity, and hence it was possible to use probes to examine CVLU. PNA FISH was performed as described previously.10 Nucleotide accession numbers GenBank accession numbers for 16S rRNA gene consensus sequences determined in this study are EU931393-EU931450. RESULTS Culture analysis Culture analysis of the 14 wounds (A–N) showed the presence of more than one species in all but one of the wounds (Tables 2 and 3). Although a diversity of other bacteria were isolated, S. aureus was detected in 13 wounds, P. aeruginosa in six, Klebsiella oxytoca in three, and Enterococcus sp. in three wounds. No obligate anaerobic species were detected in any of the wounds. Table 2. A condensed overview of the bacteria found in wound A–F1 Species Clone lib. 1 A B C D E F Staphylococcus aureus + S, C, 220 ± 6% S, ND C, ND S, C S,C,* S,C,* Pseudomonas aeruginosa + ND ND C, ND C,* C,* Staphylococcus sp. + S S S,C S,C Stenotrophomonas maltophilia + S Alcaligenes sp. + S Enterococcus sp. + C Enterococcus faecalis + C S Actinobaclulum schaalii + S Helcococcus kunzii + S Finegoldia magna + S Staphylococcus cohnnii S Corynebacterium amycolatum S Achromobacter xylosoxidans S Unidentified Gram-negative rod C Proteus sp. C Morganella morganii C Klebsiella oxytoca C Enterobacter cloacae C Peptoniphilus sp. S Uncultured Clostridia S Uncultured Clostridia S Uncultured Porphyromonas S Uncultured Bacterium S 1 Bacteria identified from wounds A–F using culture-based methods (C) and sequencing of DGGE bands (S). Quantitative PCR data are presented for S. aureus and P. aeruginosa (copies/ng DNA ± standard error of the mean, n=3). * The spatial orientation of bacteria was examined in wound D, E, and F revealing a diverse microbiota in wound E and F. Data for these two wounds are described in Table 5. Sequences also found in Clone library 1 are indicated with "+". ND, not detected. Table 3. A condesed overview of the bacteria found in wounds G–N* Species Clonelib. 2 Wounds G H I J K L M N Staphylococcus aureus + S, C, ND S, C, 120± 14% S, C, 5600± 13% S, C, NT S, C, NT S, NT S, C, 100± 5% C, ND Pseudomonas aeruginosa + C, 1400± 18% C, ND ND NT S, C, NT NT ND ND Alcaligenes sp. + S Proteus mirabillis + C Alcaligenes faecalis + C Enterococcus sp. + C Coagulase negative staphylococci + C C C Staphylococcus epidermidis S Peptoniphilus harei S S Finegoldia magna S S S Fusobacerium equinum S Peptostreptococcus anaerobius S Peptoniphilus asaccharolyticus S S S Uncultured Clostridia S Anaerococcus vaginalis S S Peptostreptococcus micros S Corynebacterium sp. S C Brevibacterium casei S Gram-negative rod C C Morganella morganii C Escherichia coli-like rod C Hemolytic Streptococcus C C Klebsiella-like rod C Klebsiella oxytoca C Bacillus sp. C Enterobacter cloacae C * Bacteria identified from wounds G–N using culture-based methods (C) and sequencing of DGGE bands (S). Quantitative PCR data are presented for S. aureus and P. aeruginosa (copies/ng DNA ± standard error of the mean, n=3). Sequences also found in Clone library 2 are indicated with "+". ND, not detected, NT, not tested. DGGE fingerprinting The results of DGGE fingerprinting are shown in Tables 2 and 3, indicated by an "S." DGGE detected S. aureus in all of the wounds except wound C, despite S. aureus being detected by the culture methods. Wound E and F showed the presence of additional uncultured bacteria. DGGE showed that the wounds also contained a variety of anaerobic bacteria with multiple findings of species such as Finegoldia magna, Anaerococcus vaginalis, Peptoniphilus asaccharolyticus, Peptoniphlus harei, and Peptostreptococcus anaerobius, often with several of these species in the same wound. P. aeruginosa was detected in only one wound with DGGE fingerprinting despite its detection in six wounds using the culture methods. An average of 3.2 species per wound were detected using DGGE fingerprinting and 3.0 species per wound were detected using culture methods. In combination, DGGE and culture identified 5.4 species per wound. Clone library and sequence analysis To elucidate the bacterial diversity in the samples, clone libraries were constructed where the amplified 16S rRNA genes were inserted into cloning vectors, thereby a separation of the different fragments and its subsequent sequencing were possible. The sequences from the two clone libraries (clone library 1 from wounds A–F and clone library 2 from wounds G–N) were divided into OTUs using a similarity level of >97%. A total of 60 clones were sequenced for clone library 1 and 94 clones for library 2. Table 4 shows the name and accession number of the closest relative for each OTU as identified by the phylogenetic analysis. Table 4. Closest relatives of the bacterial OTUs in clone libraries OTU Number* Species (BLAST) Acc. number Similarity (%) Clone library 1 1 [8/28] Staphylococcus aureus BX571856 97.1–100 2 [2/6] Alcaligenes sp. AY331576 99–100 3 [2/4] Anaerococcus sp. AM176522 99 4 [4/4] Stenotrophomonas sp. AM402950 99–100 5 [1/3] Uncultured Porphyromonas DQ130022 99–100 6 [2/3] Enterococcus faecalis DQ239694 99–100 7 [1/2] Pseudomonas aeruginosa EF064786 99–99.6 8 [1/1] Anaerococcus vaginalis AF542229 98 9 [0/1] Uncultured Anaerococcus DQ029049 95 10 [1/1] Enterobacter sp. EF088367 99 11 [1/1] Bacteroides tectus AB200228 99 12 [1/1] Actinobaculum schalli AF487680 98 13 [1/1] Helcococcus kunzii X69837 97 14 [1/1] Finegoldia magna AB109772 99 Total 57 Clone library 2 1 [7/46] Staphylococcus aureus DQ997837 98.8–99.9 2 [6/14] Pseudomonas sp. AY914070 98.7–99.0 3 [3/3] Uncultured bacterium EF511972 99.7–99.9 4 [1/3] Fusobacterium gonidoformans M58679 98.6–99.8 5 [2/2] Enterococcus faecalis DQ239694 99.8–100 6 [2/2] Acinetobacter junii AB101444 99.9 7 [1/2] Proteus mirabilis AF008582 98.6–99.8 8 [1/1] Actinobaculum schaalii AY957507 98.4 9 [1/1] Alcaligenes faecalis AY548384 97.2 10 [1/1] Helcococcus kunzii X69837 96.7 11 [1/1] Uncultured bacterium AM697030 98.2 12 [1/1] Uncultured Clostridia AY383733 99.7 Total 77 Clone library 1 showed many S. aureus and some Alcaligenes sp., Anaerococcus sp., Stenotrophomonas sp., Enterococcus faecalis, and P. aeruginosa. Clone library 2 showed a large amount of S. aureus and P. aeruginosa. Almost all OTUs have a similarity of >97% with their closest relatives. Only OTU 9 (uncultured Anaerococcus) in clone library 1 and OTU 10 Helcococcus kunzii in clone library 2 had a smaller similarity than 97% indicating that these OTUs had a lower phylogenetic resolution. The coverage ratio for the clone library 1 was 87.7% and for clone library 2 was 93.5%. The consensus sequences in clone library 1 and 2 were used to produce phylogenetic trees to determine the detailed phylogenetic relationship of the 16S rRNA gene of the clones. A neighbor joining tree, a maximum parsimony tree, and a maximum likelihood tree all showed congruent phylogenetic relationships, and the maximum likelihood tree is shown in Figure 1. The locations on the tree confirm the BLAST identification of the sequences. The sequences are distributed into five phyla: Proteobacteria, Firmicutes, Bacteroidetes, Fusobacteria, and Actinobacteria. Similar bacteria were identified in the two clone libraries, although clone library 1 did not detect any bacteria from the phylum Fusobacteria and clone library 2 did not detect any Bacteriodetes. The clone libraries were dominated by sequences related to S. aureus and P. aeruginosa, but also contained many sequences from E. faecalis, Alcaligenes faecalis, and Stenotrophomonas maltophilia. Figure 1Open in figure viewerPowerPoint Maximum likelihood (AxML) tree of consensus sequences (1364 nt compared) of consensus sequences from clone library 1 (CON#) and 2 (CONR#). The scale bar represents a 10% deviation of sequence. All 110 partial 16S rRNA gene sequences obtained from DGGE were added to the consensus maximum likelihood tree (data not shown) to confirm the result of the BLAST search. While the BLAST result was confirmed for most of the sequences, the phylogenetic analysis showed that it was not possible to distinguish the sequences identified as different Alcaligenes and Ahcromobacter species and no Peptoniphilus could be differentiated to more than the genus level. It also showed that the DGGE fingerprinting sequences most related to Fusobacterium equinum according to the BLAST were located closer to Finegoldia gonidiaformans on the tree. F. gonidiaformans was also found in clone library 2. Quantitative PCR The abundance of S. aureus and P. aeruginosa was found to vary considerably between the different wounds (Tables 2 and 3). While S. aureus could be detected by DGGE and by culturing in most samples, they were only above the limit of detection using the qPCR approach in four of the 14 ulcers investigated. P. aeruginosa could be quantified in three of the ulcers investigated. Spatial location To determine whether the bacterial composition varied throughout the wound, three wounds (D–F) were each divided in five parts and DNA was extracted from each of them. Each part was separately examined by DGGE fingerprinting and by subsequent sequencing of bands (Table 5). In wound D, only S. aureus could be detected by DGGE fingerprinting and it was present in all examined parts of the wound (data are not included in Table 5). Wound E was dominated by the aerobe S. aureus, the facultative aerobe E. faecalis, and the two anaerob