Research Article

Biology and Medicine, Vol 3 (2) Special Issue: 147-157, 2011

www.biolmedonline.com

A study of biofilm production by gram-negative organisms isolated from diabetic foot ulcer patients
*Zubair M1, Malik A1, Ahmad J2, Rizvi M1, Farooqui KJ2, Rizvi MWDepartment of Microbiology; Center for Diabetes and Endocrinology, Faculty of Medicine, J. N. Medical College and Hospital, Aligarh Muslim University, Aligarh 202002. India.
*Corresponding Author: mohammad_zubair@yahoo.co.in Abstract The present study was undertaken to study the difference in antibiotic resistance profile and minimum antibiotic concentration (MIC) of biofilm producing and non-biofilm producing gram-negative bacilli isolated from diabetic foot ulcer (DFU) patients in a tertiary care hospital in North India. Among the diabetic foot patients, 73.6% were males and 15% were females. 77.1% had T2DM whereas only 24.4% patients had T1DM. Poor glycemic control and poor HbA1c (>8) was observed in 68.7% and 70.1% patients respectively. Among the 57 patients, 97 gramnegative bacilli were isolated in which mixed bacterial infection was found in 67.8% and monomicrobial in 32.2% only. Escherichia coli was the most common (42.2%) isolate followed by Pseudomonas aeruginosa (23.7%), Klebsiella oxytoca (11.3%), Klebsiella pneumonia (9.2%), Proteus vulgaris (5.1%), Acinetobacter sp (5.1%), Proteus mirabilis (2%) and Morganella morganii (1.0%). 77.1% DFU patients had infection by biofilm producing organisms. BFP positive status was associated with the presence of neuropathy (O.R. 7.65), osteomyelitis (O.R. 2 3.14), duration of ulcer (O.R. 25.7), grade of ulcer (O.R. 9.12), necrotising ulcer (O.R. 14.4) and ulcer size >4cm (O.R. 3.30) but not with patients characteristic, type of diabetes and type of diabetes, or duration of hospital stay. Poor glycemic control in 56.1% patients, amputation (24.5%), hospital stay (38.5%) and age distribution were independently associated with risk of biofilm producing infection in diabetic foot patients. Keywords: Diabetic foot ulcer; bacterial profile; antibiotic resistance; biofilm production.
Introduction Toole et al. (2005) who observed that, the bacteria are not free floating but grow upon submerged surfaces. The basic architecture of biofilms shows that the microcolony is actually the basic structural unit of the biofilm. The exhaustive structural analysis of hundreds of monospecies in vitro biofilms, and of dozens of multispecies natural biofilms, has shown that microcolonies are discrete matrixenclosed communities of bacterial cells that may include cells of one or of many species. Depending on the species involved, the microcolony may be composed of 1025% cells and 7590% extracellular exopolysaccharide matrix (EEM). The matrix material often appears to be most dense in the area closest to the core of the microcolony, which is characterized by their lack of Brownian movement. Costerton et al. (1999) showed the arrangement of micro-colonies are in horizontal array in thin biofilms, but also form a vertical arrays in very thick sessile communities. Biofilm EEM, which is also referred to as slime (although not everything described as slime is a biofilm), is a polymeric conglomeration generally composed of extracellular DNA, proteins, polysaccharides, adhesins (PS/A) and autolysin (encoded by atIE gene) are involved in regulation of biofilm
production present in various configurations. The ica gene codes for intracellular adhesion (ICA) and may also code for PS/A and, is required for biofilm production (Toole et al., 2005; Donlan et al., 2002; Carol et al., 2005). Biofilm which forms on living or nonliving surfaces establishes a protective environment of microbial life in natural, industrial and hospital settings (Stoodley et al., 2004), which are, physiologically distinct from planktonic cells of the same organism, which, by contrast, are single-cells that may float or swim in a liquid medium (Karatan et al., 2009; Hoffman et al., 2005). When a cell switches to the biofilm mode of growth, it undergoes a phenotypic shift in behavior in which large suites of genes are differentially regulated (An et al., 2007). Biofilms are also often the site for quorum sensing influence the availability of key nutrients for biofilm formation, chemotaxis towards surface, motility of bacteria, surface adhesion and presence of surfactants are certain factors which influence biofilm formation (Carol et al., 2005; Thomas et al., 2007). According to a recent public announcement from National Institute of Health (NIH), more than 60% of all infections are caused by biofilm (Kim et al., 2001). Moreover, these ulcers adversely influence the patients quality of life, leading to decrease in

147 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
monofilament at 2 of 10 standardized plantar sites on either foot). Ulcers were assessed for signs of infection (swelling, exudates, surrounding, cellulitis, odor, tissue necrosis and crepitation) and size was determined by multiplying the longest and widest diameters expressed in cm2. Each patient was included only once in the study. All cases were monitored until discharged from the hospital. All the subjects gave informed consent and clearance was obtained from the hospital ethics committee. Microbiological Methods The microbiological methods described by Gadepalli et al. (2006) as adopted in our previous studies (Zubair et al., 2010b, c) were used. Total transfer time to the laboratory was not more than 30 minutes. Direct microscopic examination of ulcer sample was performed and all the bacterial isolates were identified to the species level using standard identification techniques (Collee et al., 1996). Susceptibility Testing Antimicrobial susceptibility testing was performed as described by the CLSI and adopted by us elsewhere (Zubair et al., 2010b,c). Antimicrobial discs used were Aztreonam (30g), Imipenem (10g), Amoxyclav (30g), Cefpodoxime (10g), Cefepime (30g), Cefoperazone (75g), Cefoperazone/sulbactam (75/10g), Cefixime (5g), Piperacillin (100g), Ceftazidime (30g), Piperacillin/tazobactam (100/10g), Ceftazidime/clavulanic acid (30/10g), Amoxycillin (20g), Cephotaxime (30g), Cephotaxime/clavulanic acid (30/10g), Ceftriaxone (30g), Cephoxitin (30g), Amikacin (30g), Chloramphenicol (30g), Gentamicin (10g), Gatifloxacin (5g), Ofloxacin (5g), Levofloxacin (5g). All discs were obtained from Hi-Media Laboratory, Mumbai, India. Interpretative criteria for each antimicrobial tested were those recommended by manufacturers guidelines (Hi-Media Labs, Mumbai, India). Biofilm Assay - Tissue Culture Plate (TCP) method The biofilm assay described by Mathur et al. (2006) was adopted. Stated briefly, 10 ml of trypticase soy broth (TSB) with 1% glucose was inoculated with a loopful of test organism from overnight culture on nutrient agar. The TSB broth was incubated at 37oC for 24 hours. The culture was further diluted 1:100 with fresh medium and flat bottom tissue culture plates (96 wells) were filled with 200l of diluted cultures individually. Uninoculated

social, physical and physiological functions (Raiber et al., 1998). Various factors including defects in host defense mechanisms (impaired leukocyte functions) are responsible for this increase in infection rates. Wound infection is known to impair wound healing in both acute and chronic DFUs (Robson et al., 1997). That most of the infections in DFU are polymicrobial in nature have recently been documented in our studies also (Zubair et al., 2010a,b). Although the numbers and type of bacteria in a wound are critical for infection to occur, recently a new concept of bacterial biofilms has emerged as a potential way to better understand how bacteria deter healing. Therefore, a better understanding of bacterial biofilms is needed, and this may ultimately result in development of novel therapeutics for the prevention and treatment of DFU infections. The biofilm producing organisms have an inherent resistance to antibiotics and in the long run they may be very damaging because of the development of immune complex diseases (Donlan et al., 2002; Raad et al., 1995; Souli et al., 1998). There are only scarce reports on biofilm formation by clinical isolates from DFU especially in North India. Keeping this in mind, the present study was undertaken to study the difference in their antibiotic resistance profile and minimum antibiotic concentration of biofilm producing and non-biofilm producing gram-negative bacilli isolated from diabetic foot ulcer in a tertiary care hospital in North India. Materials and Methods Study Design The study was carried out prospectively at the Diabetic and Endocrinology ward, J.N. Medical College, Aligarh Muslim University. Aligarh, India, from June 2009 to February 2010. Subjects studied were all in-patients of the male and female ward who had ulcer/infection in their foot with gram-negative bacterial infection. Clinical Examination A detailed clinical history and physical examination was carried out for every subject, which include a record of age, sex, anthropometric measurements, duration of ulcer, duration of diabetes and glycemic control. Foot ulcers were categorized into six grades (0-5) based on Meggit Wagner Classification System (Wagner et al, 1981). Neuropathy was quantified in each patient assessing vibration sensation using a 128 Hz tuning fork and a 10g monofilament (absence of perception of the Semmes Weinstein
148 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
qualitative variables were expressed as percentage (%). Continuous variables were compared using 2 sample t tests for independent samples. Odds ratios and 95% confidence interval (CI) were reported for independent variables associated with the outcome variable: presence of anaerobic infection. Results Clinical Males were predominant 42(73.6%) in the study subjects. Majority 44(77.1%) of subjects had T2DM. The mean age of the subjects was 49.112.4 years. The mean duration of diabetes was 12.66.4 years. Thirty-four patients (59.4%) had neuropathy, 35(61.4%) nephropathy, 32(56.1%) retinopathy, and 33(57.8%) were hypertensive. Osteomyelitis was present in 18(31.5%) subjects. Majority (77.0%) of the DFU patients were from Meggit Wagner grade II to grade IV. Grade I ulcer was found in 8.7%, Grade II in 14%, Grade III in 28%, Grade IV in 35%, and Grade V in 8.7% of patients. Majority of the subjects 31(54.3%) had lesions for >1 month before presentation at the hospital. The ulcer was necrotic in 25(43.8%) cases. Glycemic control was poor in 67(65.6%). HbA1c was <7% in 12 patients (21%), 7%-8% in 5(8.7%) and >8% in 40(70.1%) subjects. More than 38(66.6%) received surgical treatment, mainly in the form of debridement. 19(33.3%) patients were subject to amputation and 3(5.3%) died during the hospital stay (mean hospital stay 19.612.5) (Table 1). Majority of the ulcers were found on interdigits and the plantar surface (47.3% each), followed by heels (42.1%), margins (28%), malleoli (24.5%), and legs (8.7%) and on multiple (2 sites) 47.3%. Size of ulcer 4cm2 was observed in 21% 2 patients and 4cm in 64.9% patients. Microbiological Observations A total of 97 gram-negative bacteria were isolated from 57 DFU patients, averaging 1.7 species per patient. Monomicrobial infection was observed in 32.2% patients whereas polymicrobial etiology was observed in 67.8% patients. In the direct microscopic examination of ulcer samples, 96% findings correspond with the culture growth on next day and in 4% patients, direct smear result differed with their culture growth. The frequency of bacterial isolates from DFU is shown in Table 2. Escherichia coli was the most common isolate, accounting for 41(42.2%), followed by Pseudomonas aeruginosa 23(23.7%), Klebsiella oxytoca 11(11.3%), Klebsiella

sterile broth served as blank. Similarly, control organisms were also diluted and incubated. The culture plates were incubated at 37oC for 24 hours. After incubation, gentle tapping of the plates was done. The wells were washed with 200 l of phosphate buffer saline (pH 7.2) four times to remove free-floating bacteria. Biofilms which remained adherent to the walls and the bottoms of the wells were fixed with 2% sodium acetate and stained with 0.1% crystal violet. Excess stain was washed with deionized water and plates were dried properly. Optical densities (OD) of stained adherent biofilm were obtained with a micro ELISA auto-reader at wavelength of 570 nm. Experiments were performed in duplicate and the average of OD values of sterile medium were calculated and subtracted from all test values. Determination of Minimum Inhibitory Concentration (MIC) MIC was determined in doubling dilutions from 512 g/ml to 0.05 g/ml (CLSI). Antibiotic powders were obtained from Hi-Media Labs, Mumbai, India, except potassium clavulanate (clavulanic acid) which was procured from the Center for Diabetes and Endocrinology, A.M.U., Aligarh. Antibiotic Treatment Antibiotics were selected according to published recommendation (HartemannHeurtier et al., 2009). In mild infections amoxicillin clavulanic acid was given empirically by the oral route. However, in moderate infections intravenous route was preferred taking into consideration the likelihood of osteomyelitis. Considering that the causative agent was polymicrobial, we initiated ampicillin-sulbactam plus an aminoglycoside/quinolone or piperacillintazobactam or ceftriaxone plus metronidazole/clindamycin. In the presence of severe infections, surgical debridement and amputation were performed immediately after admission. Metronidazole (500 mg I.V. every 8 hours) was added to the drug regimen if cellulitis or gangrene was also present. The treatment was later modified in accordance with the culture results. The duration of the treatment was at least 4-6 weeks and prolonged in cases of osteomyelitis. All patients also received an intensive insulin treatment. Statistical Analysis The data was analyzed using SPSS version 17.0 for descriptive statistics. Quantitative variables were expressed as meansd while
149 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
biofilm producing infection. The size of ulcer more than 4 cm2 [O.R. 3.30, P < 0.89] was found in 64.9% patients with biofilm positive infection and in 14.0% patients having ulcer size less than 4 cm2. The neuropathy [O.R. 7.65, P < 0.003], osteomyelitis [O.R. 3.14, P <0.136], necrotising ulcer [O.R. 14.4, P< 0.002] and poor glycemic control (HbA1c : >8%)[O.R. 1.66, P<0.32] were significantly associated with biofilm producing bacterial infection. There was a significant relation between the biofilm producing bacterial growth with Wagners grading. Majority of the biofilm positive patients were from grade 4 [O.R. 9.12, P<0.001] followed by grade 3 [O.R. 2.56, P< 0.23], grade 2 [O.R. 2.27, P< 0.40] and grade 5 [O.R. 1.5, P< 0.68]. (Fig. 5). Discussion This study presents a comprehensive clinical and microbiological profile of infected diabetic foot ulcers in hospitalized patients with special reference to the study of biofilm production in the gram-negative bacterial isolates. With the rise in the prevalence of diabetes mellitus there is increasing problem of infections, especially foot infections. According to some studies, patients with diabetic foot infections account for 20% of hospital admissions (Shankar et al., 2005). India is the home for the largest number of diabetic individuals. As higher resistance is a growing problem, effort was made to study the association of different study characteristics with the presence of resistant organisms. The prevalence of diabetic foot ulcers among male subjects was found to be 73.6% against 26.3% in female i.e. a ratio of 2.3:1 which may be due to higher level of outdoor activity among males compared to females (Zubair et al., 2010b,c). With increasing duration of diabetes, there is increased risk of diabetes related complications especially chronic complications like sensory neuropathy. This study also reports a high prevalence of neuropathy (59.4%). There was a marked variation of sensory neuropathy from our earlier studies (Zubair et al., 2010b,c), which showed a slightly higher percentage (66.6% & 78.5%) of neuropathy in North India. Ako et al. (2006) in a Nigerian study, showed the increase in neuropathy to 77.8% and 56.8% in a South Indian study (Shankar et al., 2005). This marked variation in the prevalence may be due to difference in the methods used for the diagnosis of these conditions (10g monofilament or biothesiometer). In Table 1, duration of infection >1month, prior antibiotic use and ulcer size >4cm2 were independent predictors of

pneumoniae 9(9.2%), Proteus vulgaris 5(5.1%), Acinetobacter sp. 5(5.1%), Proteus mirabilis 2(2%) and Morganella morganii 1(1%). Biofilm Assay Among the 97 gram-negative bacterial isolates, 60(59.4%) were biofilm producers. A total of 80% P. vulgaris isolates were biofilm producers, followed by K. pneumoniae (77.7%), E. coli (63.4%), K. oxytoca (63.4%), Acinetobacter sp. (60%) and P. aeruginosa (52.1%). The lone isolate of M. morganii was a biofilm producer (Table 2). Antibiotic Resistance Profile of BFP and BFN Isolates The result of resistance studies are summarized in Fig. 1. High degree of antibiotic resistance was exhibited by all the BFP isolates compared with NBP. High degree of resistance by BFP isolates was observed against cefoparazone (79.6%) followed by piperacillin (68.4%), cephotaxime (67.3%), amoxyclav (64.3%), cefixime (64.3%), amoxycillin (63.3%), ofloxacin (63.3%), cefepime (59.2%), gatifloxacin (57.1%), levofloxacin (51.0%), cefpodoxime (49.0%), ceftriaxone (44.9%), ceftazidime (42.9%), amikacin and gentamicin (40.8% each), astreonam (39.8%), cephoxitin (36.7%), chloramphenicol (31.6%), imepenem (24.5%), piperacillin+tazobactum (21.4%), cefotaxime+clavulanic acid (12.2%), and Ceftazidime+clavulanic acid (9.2%). Minimum Inhibitory Concentration (MIC) The MIC values of the piperacillin (with/without tazobactam), cefoparazone (with/without sulbactam), ceftazidime (with/without clavulanic acid) and levofloxacin between the BFP and NBP were given in Table 3. Percentage of BFP isolates that had an MIC of 2g/ml was 93.3% for cefoparazone followed 90% for piperacillin, 81.6% for ceftazidime, and 75% for levofloxacin. The isolates that had an MIC 2g/ml antibiotics with inhibitor were 80% for piperacillin+tazobactum, followed by 73.3 % for cefoparazone+sulbactum and 48.3% for ceftazidime + clavulanic acid. Correlation of Biofilm Assay and Clinical Characteristics of DFU Patients Table 1 also shows the result of univariate analysis of factors to be associated with the presence of biofilm producing organism infections. The age distribution [O.R. 1.23, P = 0.489], Type 2 diabetes [O.R. 2.16, P<0.207], duration of ulcer >1 month [O.R. 25.7, P < 0.001] was observed in 52.6% patients having
150 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
antimicrobial attack (Mertz, 2003). The biofilm also provides a physical protection to bacteria because antimicrobial agents are also ineffective at penetrating the biofilm, decreasing the concentration acting on the bacterial cells within the biofilm and as a consequence their efficacy (Mah and Toole, 2001). In addition to the resistance to antimicrobials, biofilms also appear to have an antiphagocytic property within the biofilm, which renders leukocytes present within the matrix ineffective (Leid, 2002). Additionally, there appears to be a component within the polysaccharide that inactivates and traps both complement and host antibodies. These factors lead to an accumulation of host immune factors that can lead to host tissue damage and eventually chronic inflammation (Percival and Bowler, 2004). The idea of disrupting a biofilm that is already formed is attractive. This could be accomplished in a number of ways, including physical methods and/or application of topical substances. Among potential physical methods, debridement, electrical stimulation, or ultrasound could be used. Debridement may not only remove the bacteria and biofilm but also may aid in the removal of necrotic tissue for which the bacteria would thrive on. Electrical stimulation has been used over the years to assist penetration of various topical agents but have a limited application (e.g., electroporation and electrophoresis have been shown to enhance the penetration of a photosensitizer) (Johnson and Oseroff, 2002). Changing the perspective about chronic infectious disease to include biofilm enables two important insights. First, it opens new methods for detection and treatment. Second, it provides a global reconceptualization of many chronic infectious diseases as resulting from a biofilm, allowing biofilm principles to be shared across disciplines. Recent studies have investigated new methods for detecting the components of a biofilm. Several investigations have used modern molecular methods, such as denaturing gradient gel electrophoresis and denaturing high performance liquid chromatography, along with imaging techniques including fluorescent in situ hybridization. Also, molecular methods such as polymerase chain reaction (PCR) and pyrosequencing in conjunction with conventional culture methods have been used to determine the bacterial species composition of chronic infections (Dowd et al., 2008). Performing molecular tests as part of routine bacterial analysis is becoming a real option for clinical laboratories. These tests could include

infection. Thus patients with a large ulcer, with a history of prior antibiotic use and duration of infection >1month were more likely to harbor BFP organisms. In the present study, mean duration of ulcer was found to be 41.547.6 days with 54.3% having ulcer for more than 1 month. About 78.9% presented with a large ulcer of approximate size of >4cm2, thereby accounting for approximately 77.1% of the patients presenting with Wagners grade II and IV. The reasons for presentation with advanced grade and stage of ulceration could be because of lack of structured health care delivery in the country, attempted selfmedication and trust in traditional healers (Boulton et al., 2001; Zubair et al., 2010b,c). Diabetic foot infections are usually polymicrobial in nature and this has been well documented in the literature. In our study also, we found polymicrobial etiology in 67.8% and monomicrobial in 32.2% patients with the rate of isolation of about 1.7 bacteria per patient which is higher than the previous reports (Zubair et al., 2010a,b,c) whereas Gerding et al., (1995) and Gadepalli et al. (2006) have reported higher isolation rate of 2.0%-5.8%. The present study also confirms the high resistance among the DFU isolates which was extremely common in hospitalized patients with diabetic foot ulcers. This is in accordance with the reports of Hartemann-Heurtier et al. (2009) and Zubair et al. (2010a,b,c). This high degree of antibiotic resistance may be due to the fact that ours is a tertiary care hospital with widespread usage of broad spectrum antibiotics leading to selective survival advantage of pathogen. Our results of antimicrobial resistance pattern were similar to the recent studies done in India and outside (Shankar et al, 2005; Raja et al., 2007). Gramnegative bacteria that are regarded as normal flora of the skin, like P. aeruginosa, may cause severe tissue damage in diabetics and should never be automatically disregarded as insignificant in diabetic foot ulcers (Zubair et al., 2010b). Another reason for this high antimicrobial resistance among the BFP appears to be due to the close cell-cell contact that permits bacteria to more effectively transfer plasmids to one another than in the planktonic state. These plasmids can encode for resistance to several different antimicrobial agents (Mah and Toole, 2001). Another factor contributing to resistance is quorum sensing, which through the processes described above can force bacteria into a slow-growing state when placed in an environment with adverse growth conditions; when in this state of intermission, bacteria are less susceptible to

151 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
Acknowledgement The authors would like to thank Dr. Rafat Fatima, Diabetic Educator, Center for Diabetes and Endocrinology, for dietary advice and monitoring the patients diet. The authors also like to thank Dr. Danish H and Dr. S. Y. Nahid Zaidi, residents of the department of Medicine, JNMC, Aligarh Muslim University.
References Ako-Nai AK, Ikem IC, Akinloye OO, Aboderin AO, Ikem RT, Kassim OO, 2006. Characterization of bacterial isolates from diabetic foot infections in IleIfe, Southwestern Nigeria. The Foot, 16(3): 158164. An D, Parsek MR, 2007. The promise and peril of transcriptional profiling in biofilm communities. Current Opinion in Microbiology, 10(3): 2926. Boulton AJ, Vileikyte L, 2001. Diabetic foot problems and their management around the world. In Levin and O Neals The Diabetic Foot. Sixth Edition. St. Louis, MO: Mosby, 6: 261-71. Carol A, Kumamoto, Marcelo DV, 2005. Alternative Candida albicans life style: Growth on surfaces. Annual Review of Microbiology, 59: 113- 133. Christensen GD, Simpson WA, Bisno AL, Beachey EH, 1982. Adherence of slime producing strains of Staphylococcus epidermidis to smooth surfaces. Infection and Immunity, 37(1): 318-326. CLSI. Clinical and Laboratory Standards Institute, 2007. Performance Standards for Antimicrobial Susceptibility Testing: Seventeenth Informational Supplement. M100-S17, vol. 27, no. 1. Collee JG, Marr W, 1996. Mackie and McCartney Practical Microbiology, 14th edition. London: Churchill Livingstone, 77: 637650. Costerton JW, 1999. Introduction to biofilm. International Journal of Antimicrobial Agents, 11: 217221. Donlan RM, Costerton W, 2002. Biofilms: Survival mechanisms of clinically relevant microorganisms. Clinical Microbiological Review, 15(2): 167-193. Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BW, James GW, Wolcott GD, 2008. Survey of bacterial diversity in chronic wounds using Pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiology, 8: 43. Ehrlich GD, Hu FZ, Shen K, Stoodley P, Post JC, 2005. Bacterial plurality as a general mechanism driving persistence in chronic infections. Clinical Orthopaedics and Related Research, 437: 20-24.

methods such as PCR, reverse transcriptase PCR, microarrays, antigen testing, and rapid sequencing. Only a few of these methods are being used to test for certain pathogens, but culture-free identification of all pathogens and their corresponding resistance markers may soon become routine (Espy et al., 2006). A biofilm focus also provides new strategies for treatment of chronic infections. Biofilm-based treatments might block initial bacterial attachment to a surface, block or destroy EPS formation, interfere with cell-cell signalling pathways, and use bacteriostatic or bactericidal agents at the same time. Concomitant therapies that not only attempt to eradicate bacteria but also affect the biofilms community structure and communications may prove more effective than a single or sequential strategy such as antibiotic therapy (Ehrlich et al., 2005). This multimodality approach to therapy is commonly used in other areas of medicine, such as the treatment of human immunodeficiency virus for which combination antiretroviral therapy is used to achieve the best clinical outcome. Conclusion Diabetic foot infections are a significant burden on patients as well as a burden on the health care delivery system. It is important to quickly and effectively identify and treat these ulcers and prevent complications. Biofilm formation on these wounds may be responsible for the chronicity of these wounds and for their common infectious complications. The presence of biofilm also represents an important barrier to effective treatment. Although in vitro study of novel approaches to control or eradicate biofilm formation are being performed, in vivo testing is necessary because various factors (e.g., wound fluid, proteases, growth factors, and so forth) need to be taken into consideration to determine the true efficacy of these agents. Treating the DFU by shifting from the planktonic model of microbiology to the biofilm model makes available new methods for detection and treatment. Because of molecular methods, science now has the ability to detect biofilms and understand the implications of interspecies chaos that contribute to infections. With these new scientific approaches along with coordination of clinical and laboratory efforts, education, and research, it is possible to imagine overcoming much of biofilm disease.

152 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
Table 1: Demographic presentation of DFU patients in response to biofilm assay positive and negative bacterial infections (meansd and n(%) of otherwise indicated). N=57 Total Biofilm + Biofilm POR(95%CI) (n=44) (n=13) value Sex Male 42(73.6) 0.532 0.8(0.18-3.4) Female 15(26.3) Age distribution (years) 49.112.4 44.67.3 54.310.2 <40 10(17.54 7(12.2) 3(5.2) 41-60 1.23(0.35-4.3) 33(57.8) 26(45.6) 7(12.2) 0.489 >61 14(24.5) 11(19.2) 3(5.2) Type of Diabetes Type 1 14(24.4) 9(15.7) 5(8.7) Type 2 2.16(0.57-8.0) 44 (77.1) 35(61.4) 9(15.7) 0.207 Duration of Ulcer 41.5 47.5 39.62.6 22.71.0 < 1month 26 (45.6) 14(24.5) 12(21.0) >1 month 0.0001 25.7(3.0-217.7) 31(54.3) 30(52.6) 1(1.7) Hospital stay(days) 19.6 12.5 20.612.3 9.210.19(33.3) 10(17.5) 9(15.70 20-40 1.22(0.42-3.5) 24(42.1) 22(38.5) 2(3.5) 0.46 >40 14(24.5) 12(21.0) 2(3.5) Ulcer Grade (Wagner) grade 0 3(5.2) 0(0) 3(5.2) grade 1 5(8.7) 0(0) 5(8.7) grade 2 2.27(0.25-20.3) 8(14) 7(12.2) 1(1.7) 0.40 grade 3 2.56(0.5-13.1) 16(28) 14(24.5) 2(3.5) 0.23 grade 4 0.001 9.12(1.08-76.3) 20(35) 19(33.3) 1(1.7) grade 5 1.2(0.12-11.7) 5(8.7) 4(7.0) 1(1.7) 0.68 Status discharge 54 (94.7) 41(71.9) 12(22.2) Dead 0.878(0.08-9.2) 3 (5.3) 2(3.5) 1(1.7) 0.656 Treatment conservative 38(66.6) 30(52.6) 8(14.0) amputation 0.74(0.28-2.6) 19 (33.3) 14(24.5) 5(8.7) 0.447 Diabetes duration(years) 12.6 6.40 14.92.6 7.62.7 Size of ulcer 20.14 44.85 19.23.7 9.82.cm(21) 7(12.2) 5(8.7) >4 cm2 3.30(0.83-13.1) 45 (78.9) 37(64.9) 8(14.0) 0.89 Complications neuropathy 0.003 7.65(1.9-30.1) 38(66.6) 34(89.4) 4(10.5) nephropathy 35(61.4) 27(77.1) 8(22.8) 0.627 0.49(0.27-3.54) retinopathy 32(56.1) 22(68.7) 10(31.2) 0.078 0.30(0.07-1.24) hypertension 33(57.8) 24(72.7) 9(27.2) 0.269 0.53(0.14-1.99) osteomyelitis 3.14(0.61-15.9) 18(31.5) 16(88.8) 2(11.1) 0.136 Nature of Ulcer necrotising 0.002 14.4(1.72-120) 25(43.8) 24(96) 1(4) non-necrotising 32(56.1) 20(62.5) 12(37.5) Body Mass Index 20.594.41 20.32.1 18.61.8 Plasma Glucose fasting 174.2885.33 184.724.7 142.42.8 postprandial 222.7292.18 238.432.7 187.412.7 HbA1c % 10.112.50 10.71.7 7.12.5 <7 %(good control) 12(21.0) 9(15.7) 3(5.2) 7-8 % (fair control) 5(8.7) 3(5.2) 2(3.5) >8 % (poor control) 40(70.1) 1.66(0.45-6.11) 32(56.1) 8(14.0) 0.32

153 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
Table 2: Gram-negative bacilli isolated from 57 diabetic foot ulcers (N=97). Name of DFU isolates Escherichia coli Pseudomonas aeruginosa Klebsiella oxytoca Klebsiella pneumoniae Proteus vulgaris Proteus mirabilis Acinetobacter sp Morganella morganii Total Biofilm assay Positive Negative 26(63.4) 15(36.5) 12(52.1) 13(47.9) 7(63.6) 4(36.4) 7(77.7) 2(22.3) 4(80) 1(20) 2(100) 3(60) 2(40) 1(100) 60(59.4) 37(38.1) Total 41(42.2) 23(23.7) 11(11.3) 9(9.2) 5(5.1) 2(2.0) 5(5.1) 1(1.0) 97
Fig. 1: Average resistance percentage of biofilm positive and negative gram-negative DFU isolates tested against various antibiotics.

Percentage

6.1 20.4

40.0 20.0 0.0

24.5 8.2 3.1 10.2

1.09.2

Biofilm positive %

Biofilm negative %

Table 3: MIC of gram-negative bacilli (GNB) isolated from 57 DFU patients (N=97). MIC Piperacillin Piperacillin+Tazobactum Cefoparazone Cefoparazone+Sulbactum Ceftazidime Ceftazidime+Clavulanic acid Levofloxacin Biofilm producers 2g/ml 54(90) 48(80) 56(93.3) 44(73.3) 49(81.6) 29(48.3) 45(75) Non-biofilm producers 2g/ml 6(10) 12(20) 1(6.7) 16(27) 11(18.4) 31(51.6) 15(25)
MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
Fig. 2: Fasting and postprandial blood glucose level among DFU patients having infection with the biofilm producing and non-producing gram-negative bacterial infections at the time of admission and discharge from the hospital. 129

Blood glucose(mg/dl)

admission
Biofilm producer fasting Non-biofilm producer fasting

discharge

Biofilm producer postprandial Non-biofilm producer postprandial
Fig. 3: HbA1c values among the DFU patients having infection with the biofilm producing and non-producing gram-negative bacterial infections at the time of admission and discharge from the hospital. 10

8.9 8.1

7.8 6.8
Non-biofilm producer HbA1c

Biofilm producer HbA1c

Fig. 4: Tissue culture plate showing the result of biofilm assay, A1 and B1 were blank.
155 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
Fig. 5: Images of Diabetic Foot Ulcer.

Fig.5:

Images of Dc foot ulcer.
156 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied
Percival L, Bowler P, 2004 Biofilms and their potential role in wound healing. Wounds, 16: 234 240. Raja NS, 2007. Microbiology of diabetic foot infections in a teaching hospital in Malaysia: a retrospective study of 194 cases. Journal of Microbiology, Immunology and Infection, 40(1): 3944. Reiber G, Lipsky B, Gibbons G, 1998. The burden of diabetic foot ulcers. American Journal of Surgery, 176(2A): 5S10S. Robson MC, 1997. Wound infection: a failure of healing caused by an imbalance of bacteria. The Surgical Clinics of North America, 77(3): 637-50. Shankar EM, Mohan V, Premalatha G, Srinivasan RS, Usha AR, 2005. Bacterial etiology of diabetic foot infections in South India. European Journal of Internal Medicine, 16: 567-570. Souli M, Giamarellou H, 1998. Effects of slime produced by clinical isolates of coagulase negative Staphylococci on activities of various antimicrobial agents. Antimicrobial Agents and Chemotherapy, 42(4): 939-941. Stoodley L, Costerton JW, Stoodley P, 2004. Bacterial biofilms: from the natural environment to infectious diseases. Nature Reviews Microbiology, 2 (2): 95108. Thomas D, Day F, 2007. Biofilm formation by plant associated bacteria. Annual Review of Microbiology, 61: 401-422. Toole GO, Kaplan HB, Kolter R, 2000. Biofilm formation as microbial development. Annual Review of Microbiology, 54: 49-79. Wagner FW, 1981. The dysvascular foot: a system of diagnosis and treatment. Foot Ankle, 2: 64122. Zubair M, Abida M, Jamal A, 2010a. Clinicobacterial study and drug resistance profile of diabetic foot infections in North India. Diabetes, 59: supplement 1(June): A781 (Abstract: 2398-PO). Zubair M, Abida M, Jamal A, 2010b. Clinicomicrobiological study and antimicrobial drug resistance profile of diabetic foot infections in North India. The Foot (in press). Zubair M, Abida M, Jamal A, 2010c. Clinicobacteriology and risk factors for the diabetic foot infection with multidrug resistant microorganisms in North India. Biology and Medicine, 2(4): 22-34.

Espy MJ, Uhl JR, Sloan LM, Buckwalter SP, Jones MF, Vetter ME, Yao JDC, Wengenack YL, Rosenblatt YL, Cockerill FR, Smith TF, 2006. Realtime PCR in clinical microbiology: applications for routine laboratory testing. Clinical Microbiological Reviews, 19(1): 165- 256. Gadepalli R, Dhawan B, Sreenivas V, Kapil A, Ammini AC, Chaudhry R, 2006. A clinicomicrobiological study of diabetic foot ulcers in an Indian tertiary care hospital. Diabetes Care, 29: 1727-1732. Gerding DN, 1995. Foot infections in diabetic patients: the risk of anaerobes. Clinical Infectious Diseases, 20 (suppl): S2838. Hartemann-Heurtier A, Robert J, Jacqueminet S, Ha Van G, Golmard JL, Jarlier V, Grimaldi A, 2004. Diabetic foot ulcer and multidrug-resistant organisms: risk factors and impact. Diabetic Medicine, 21: 710715. Hoffman LR, D'Argenio DA, MacCoss MJ, Zhang Z, Jones RA, Miller SI, 2005. Aminoglycoside antibiotics induce bacterial biofilm formation. Nature, 436 (7054): 11715. Johnson PG, Oseroff AR, 2002. Electrically enhanced percutaneous delivery of deltaaminolevulinic acid using electric pulses and DC potential. Photochemistry and Photobiology, 75: 534540. Karatan E, Watnick P, 2009. Signals, regulatory networks, and materials that build and break bacterial biofilms. Microbiology and Molecular Biology Reviews, 73 (2): 31047. Kim L, 2001. Riddle of biofilm resistance. Antimicrobial Agents and Chemotherapy, 45(4): 999-1007. Leid J, 2002. Human leukocytes adhere to, penetrate, and respond to Staphylococcus aureus biofilms. Infection and Immunity, 70: 63396345. Mah T, OToole G, 2001. Mechanisms of biofilm resistance to antimicrobial agents. Trends in Microbiology, 9: 3439. Mathur T, Singhal S, Khan S, Upadhyay DJ, Fatma T, Rattan A, 2006. Detection of biofilm formation among the clinical isolates of Staphylococci: An evaluation of three different screening methods. Indian Journal of Medical Microbiology, 24(1): 2529. Mertz P, 2003. Cutaneous biofilms: friend or foe? Wounds, 15: 129132. National Committee for Clinical Laboratory Standards, 2002. Performance Standards for Antimicrobial Susceptibility Testing: Twelfth Informational Standard. M100-S12, vol. 22, no. 1.
157 MAASCON-1 (Oct 23-24, 2010): Frontiers in Life Sciences: Basic and Applied

research

Bacterial diversity in surgical site infections: not just aerobic cocci any more
Objective: To evaluate the microbial diversity in chronic surgical site infections (SSIs). Method: Bacterial populations in 23 chronic SSIs were identied using bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP), which is an universal bacterial identication method. These results were then validated using quantitative polymerase chain reaction (qPCR). Results: bTEFAP identied two previously uncharacterised Bacteroidales in all of the SSIs and showed that it was the predominant population in the majority of these chronic wounds. Other bacteria identied included Corynebacterium spp., Peptoniphilus spp., Staphylococcus spp., Staphylococcus aureus, Serratia marcescens, Prevotella spp. and Pseudomonas aeruginosa. Rarefaction analysis of the data indicated that, on average, six genera occurred in any given SSI, suggesting that such infections are multispecies. On average, over 60% of the bacteria evaluated in the SSIs were anaerobic bacilli. The previous literature indicates that aerobic cocci predominate in such wounds. Conclusion: This modern molecular survey indicates that our previous understanding of which bacteria cause SSIs may be faulty. The high prevalence of anaerobic bacilli and the overwhelming predominance of two previously uncharacterised Bacteroidales suggest that such bacteria may be a leading contributor to such infections. Further research on the identication and treatment of such bacteria are warranted. Declaration of interest: Scott E. Dowd is director of a clinical molecular diagnostic company.
surgical site infection; biofilm; Staohylococcus aureus; bTEFAP; qPCR
t was generally thought that Gram-positive aerobic cocci, particularly Staphylococci, are the primary cause of surgical site infections (SSIs).1,2 Indeed, the isolates most commonly identied in SSIs using culture-based methods are Staphylococcus aureus, coagulase-negative staphylococcus, Enterococcus spp. and Escherichia coli.1-3 However, this does not take account of the fact that the biolm phenotype, with its multispecies communities, is the natural state of existence for most types of bacterium, and that over 99% of bacteria identied in every environment are organised in biolm communities.4 Indeed, it is now becoming accepted that not only are biolms the prevalent cause of chronic wound infection,5 but also that culture methods cannot identify biolms.5-7 SSIs have many of the characteristics of chronic wounds specically and of chronic infections in general.5,6 Their management therefore requires an understanding of biolm phenotype bacteria.8 Few studies have used modern molecular methods to evaluate the microbial diversity in SSIs. Previous studies have relied on clinical and laboratory culture methods to evaluate which organisms can be isolated in pure culture from an SSI. As Staphylococcus, Streptococcus, Escherichia and Pseudomonas
J O U R N A L O F WO U N D C A R E V O L , N O 8 , A U G U S T 0 9
spp. grow easily in clinical and laboratory culture media and are easily isolated, it is not surprising that they are the microorganisms most frequently associated with SSIs. However, throughout the literature it is stated that, on average, fewer than 5% of all bacteria can be easily grown in laboratory culture media. Therefore, 95% of all bacteria that might be associated with or causative factors for SSIs are never isolated and so have never been identied. This limits our ability to treat such infections. Molecular techniques have an advantage over culture methods as they do not rely on growing limited species of bacteria in the laboratory, but instead are able to identify all microorganisms contained within a sample on the basis of their genetic material.9-12 Examples of molecular techniques that pose an alternative to traditional bacteriological analysis are bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) and quantitative polymerase chain reaction (qPCR). bTEFAP is a molecular technology that uses genetic information to identify all bacteria contained within a wound sample. qPCR is another mo lecular detection technique that enables us to detect the specic genetic signatures of bacteria and provide relative or even absolute quantication of that genetic material.

R.D.Wolcott, MD, Medical Director, Southwest Regional Wound Care Centre, Lubbock, Texas, US; V. Gontcharova,1 MS, Bioinformatician; Y. Sun,1 PhD, Molecular Microbiologist; A. Zischakau,1 MS, CLSp(MB), Molecular Diagnostic Technician; S.E. Dowd,1 PhD, Director of Medical Biolm Research; 1 Research and Testing Laboratory, Lubbock, Texas, US. Email: sdowd@ pathogenresearch.org

Method

Table 1. Patient demographics and wound characteristics
Sample code 1 1b 5b 11b 14 14b 19 Sex Female Female Female Female Male Male Male Female Female Male Male Male Male Male Female Female Male Male Male Male Male Male Female Age 41 Diabetes No No Yes No No No No Yes No No No No No No No No No No No No No No No Location Left lateral ankle Left lateral ankle Left biceps Right ankle Right forearm Right hip Right hip Right below-knee amputation Right elbow Left plantar rst metatarsal Abdomen Right shoulder Abdomen Abdomen Right hernia Mid back Right knee Right knee Abdomen Left heel S/P Harrington Rod back Chest Abdominal Duration (months) 22 1.12 10
Debridement samples from 23 separate SSIs were collected from patients at the Southwest Regional Wound Care Center (Lubbock, Texas) in accordance with Western Institutional Review Board protocol number 20062347. The debridement samples were collected with sterile tools after cleansing the wound using sterile techniques. All samples were frozen in sterile collection tubes at -80oC until DNA extraction was performed. Table 1 provides a summary of these 23 SSIs. The inclusion criterion was wounds classied as SSIs of at least one months duration that had been referred to Southwest Regional Wound Care Clinic.

DNA extraction

After thawing, portions of the debridement samples (mean 200mg 100) were recovered using sterile forceps. The samples were placed in 2ml sterile micro centrifuge tubes, centrifuged at 14,000 revolutions per minute (rpm) for 30 seconds and resuspended in 500l RLT buffer (Qiagen, CA, US) with -mercaptoethanol. A sterile 5mm steel bead and 500l 0.1mm sterile glass beads (Scientic Industries, NY, USA) were added to achieve complete bacterial lysis in a Qiagen TissueLyser, run at 30Hz for ve minutes. Samples were centrifuged briey and 100l 100% ethanol was added to a 100l aliquot of the sample supernatant. This mixture was then added to a DNA spin column, and DNA recovery protocols were followed as instructed in the QIAamp DNA Mini Kit (Qiagen). DNA was eluted from the column with 30l water, and samples were diluted accordingly to a nal concentration of 20ng/l for use with all qPCR reactions. DNA samples were quantied using a Nanodrop spectrophotometer (Nyxor Biotech, Paris, France).

bTEFAP

bTEFAP was performed as described previously.9-11 bTEFAP sequence processing pipeline Once the DNA sequence data for each sample are generated, the goal is to take the raw genetic information and manipulate it to determine the bacteria present within each sample and their relative percentages. Processing removes poor-quality or noisy genetic data, organises it and then uses a computer algorithm to identify the microorganisms from which the data were derived. Custom software written in C# within a Microsoft.NET development environment was used for all post-sequencing processing. In summary, qualitytrimmed sequences were derived directly from FLX sequencing run output les. Tags (articially created genetic signatures used to identify which sample or specimen the sequence was derived from) were extracted from the FASTA les (which contained all

This study used bTEFAP to identify all microorganisms found in 23 wounds with chronic SSIs. It then validated these results by using a second independent measure qPCR. The combined use of these techniques enabled us to evaluate the diversity and predominance of all microorganisms identied in each sample. We anticipate, therefore, that this study will shed light on the diversity of SSIrelated microbial communities and demonstrate that a reliance on clinical culture has limited our knowledge of this.
Fig 1. Diversity and distribution of species among the SSIs
Bacteroidales unknown A Corynebacterium spp. Bacteroidales unknown B Staphylococcus spp. Staphylococcus aureus Bacteroides fragilis Serratia marcescens Finegoldia spp. Dermabacter hominis Staphylococcus haemolyticus Anaerococcus octavius Propionibacterium acnes Granulicatella adiacens Escherichia coli Streptococcus agalactiae Propionibacterium granulosum Alistipes negoldia Anaerobiospirillum succiniciproduce Prevotella spp. Peptoniphilus spp. Pseudomonas aeruginosa
8 11b 17 1b 5 14b 13 5b 12
Percentage of bacteria: colour key % 100.00 22.13 4.90 1.08 0.00

Bacterial species

Sample no.
of the sequence information) into individual samplespecic les, based on the tag sequence. Tags that did not have 100% homology to the sample designation were not considered. Sequences less than 150bp after quality trimming were not considered. The resulting sequence information was parsed among the 23 samples, averaging at least 1,000 sequence reads per sample. These were then depleted of chimeric (contaminated) sequences and evaluated using Basic Local Alignment Search Tool (BLAST) against a custom database derived from GenBank (http://ncbi.nlm.nih.gov). A post-processing algorithm generated a summary of the genetic search-engine data for each sample. Bacteria classied at the species level had a sequence identity (divergence) greater than 96.5%; at the genus level, it was between 94% and 96.5%; and at the closest family level, it was less than these values. No sequences were analysed with less than 84% similarity from known sequences. Following best-hit processing, a secondary postprocessing algorithm was used to compile data from each sample and determine the relative predicted ratios of each organism within each sample.

The diagnostic panel described previously was used.13 Each sample was screened to determine the relative percentage of S. aureus, Pseudomonas aeruginosa, Serratia marcescens and E. coli, and validated by comparing this with the results obtained using bTEFAP. Two unknown Bacteriodales were identied using the above techniques. One of these, unknown Bacteriodales A, was further characterised and validated using SYBR green qPCR. Using full-length DNA sequences obtained from the sample, quantitative primers were designed using a custom computer program. To validate and conrm the percentage of the unknown Bacteroidales A, the results were compared with an universal eubacterial quantitative SYBR green specic primer set. This allowed us to detect this unknown bacterium and determine that the original sequencing information is reproducible.

Statistics

Statistics were performed using the comparative functions and multivariate hierarchical clustering methods of NCSS 2007 (Kaysville, UT, USA). Rarefaction analysis was performed as described previously.11
Table 2. Summary of the primary genera identied in the SSIs
Name Bacteroides Staphylococcus Prevotella Corynebacterium Peptoniphilus Escherichia Serratia Finegoldia Propionibacterium Anaerobiospirillum Pseudomonas Streptococcus Dermabacter Anaerococcus No. of samples Mean % 45.4 29.9 2.9 41.5 9.7 1.0 4.8 9.4 2.7 0.5 31.6 4.2 10.7 12.9 SD 42.1 41.1 5.8 38.2 21.9 1.4 6.4 16.4 3.1 0.4 48.1 5.4 17.4 16.6 Maximum % 100 99.98.8 54.5 3.8 17.2 38.2 7.6 1.1 87.0 10.5 30.8 24.7 Gram stain* + + + + + + + + Oxygen tolerance Anaerobic Facultative anaerobe Anaerobic Aerobic Anaerobic Facultative anaerobe Facultative anaerobe Anaerobic Anaerobic Anaerobic Aerobic Facultative anaerobe Aerobic Anaerobic Morphology (shape) Rod Cocci Rod Rod Cocci Rod Rod Cocci Rod Sprial Rod Cocci Rod Cocci
No. of samples = the no. of samples in which the genera were identied Mean % = the average percentage of that genera in each of the samples SD = standard deviation of these percentages Maximum % = the maximum percentage of the genera in each sample * + Gram-positive; - Gram-negative Anaerobic bacteria are unable to propagate in laboratory media in the presence of oxygen; facultative anaerobes can grow both in the presence of oxygen and without it; aerobic bacteria can grow in the presence of oxygen

Results

The bTEFAP results are shown in Fig 1, which provides an overview of the diversity data and depicts the distribution of bacterial species among the SSIs. The colours provide a relative indication of the prevalence (percentage) of each species within a given SSI (sample x-axis). Red indicates a high percentage, while black indicates absence of a given genus/species. The scale on the left provides the key to the heat map colours. Fig 1 indicates that: Bacteroidales unknown A isolate occurred in all but one of the SSIs, and was the predominant population in seven samples Bacteroidales unknown B was predominant in three samples One of the two unknown Bacteroidales (A and B) occurred in all of the SSIs and was the primary population in 11 of the 23 SSIs Corynebacterium spp. were the predominant populations in seven of the samples and occurred in 12 Staphylococcus spp. (probably representing a new species of Staphlococcus) was predominant in two of the samples

S. aureus occurred in nine samples but was predominant in only one. Other bacteria that occurred frequently were S. marcescens, E. coli, Anaerobiospirillum succiniciproduce, Prevotella and Peptoniphilus spp. P. aeruginosa occurred in three samples but was predominant in only one. Table 2 provides a summary of the primary genera identied in these SSIs. The two species most frequently identied were Bacteroides and Staphyloccocus (primarily S. aureus). The predominance of anaerobes such as Peptoniphilus, Finegoldia and Anaerococcus was also notable. This is supported in Fig 2, which shows that anaerobic bacilli predominated in the SSI samples. Further summary of Fig 2 illustrates that the anaerobes were the predominant populations in well over half the samples, with facultative aerobes occurring as predominant populations in only nine. The predominance of bacilli and anaerobes contradicts previous reports that aerobic cocci are the primary contributor to SSI.9,11 The most ubiquitous
References 1 Owens, C.D., Stoessel, K. Surgical site infections: epidemiology, microbiology and prevention. J Hosp Infect 2008; 70: S2, 3-10. 2 Tourmousoglou, C.E., Yiannakopoulou, E.C., Kalapothaki,V. et al. Surgical-site infection surveillance in general surgery: a critical issue. J Chemother 2008; 20: 3, 312-Shukla, S., Nixon, M., Acharya, M. et al. Incidence of MRSA surgical-site infection in MRSA carriers in an orthopaedic trauma unit. J Bone Joint Surg Br 2009; 91: 2, 225-228. 4 Ehrlich, G.D., Hu, F.Z., Shen, K. et al. Bacterial plurality as a general mechanism driving persistence in chronic infections. Clin Orthop Relat Res 2005; 437, 20-24. 5 Wolcott, R.D., Ehrlich, G.D. Biolms and chronic infections. JAMA 2008; 11: 299 (22), 2682-2684. 6 Wolcott, R.D., Rhoads, D. D., Dowd, S.E. Biolms and chronic wound inammation. J Wound Care 2008; 17: 8, 333-341. 7 Wolcott, R.D., Rhoads, D. D. A study of biolm-based wound management in subjects with critical limb ischaemia. J Wound Care 2008; 17: 4, 145-154. 8 Soderquist, B. Surgical site infections in cardiac surgery: microbiology. APMIS 2007;115: 9, 1008-1011. 9 Dowd, S.E., Wolcott, R.D., Sun,Y. et al. Polymicrobial nature of chronic diabetic foot ulcer biolm infections determined using bacterial tag encoded FLX amplicon pyrosequencing (bTEFAP). PLoS ONE 2008; 3: 10, e3326. 10 Dowd, S.E., Sun,Y., Wolcott, R.D. et al. Bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: bacterial diversity in the ileum of newly weaned Salmonella-infected pigs. Foodborne Pathog Dis 2008; 5: 4, 459-472. 11 Dowd, S.E., Callaway, T. R., Wolcott, R.D. et al. Evaluation of the bacterial diversity in the feces of cattle using 16S rDNA bacterial tag-encoded FLX amplicon pyrosequencing (bTEFAP). BMC Microbiol 2008; 8: 125.

Fig 2. Distribution and contribution of major phenotypic bacterial characteristics from surgical site infections
Percentage of bacteria: colour key % 100.00 22.13 4.90 Phenotype Coccus 1.08 0.00 Facultative

Bacillus

Anaerobe

Motile

Aerobes
10 14b 13 5b 1b 6 11b 12 4
organisms were two previously uncharacterised bacteroidales, which occurred in all but two of the SSIs, suggesting that these unknown anaerobes may be major contributors to SSI infections. The novel Bacteroidales organism is closely related to Bacteroides spp. The samples were all unrelated; the sampling dates ranged over a 12-month period and there was no correlation between the sampling date and the occurrence of these organisms (data not shown). Results of the qPCR correlated highly with the bTEFAP results (85%, p<0.01), indicating that the bTEFAP results were accurate. Finally, we evaluated the diversity of bacteria among the wounds. Rarefaction analysis, which was performed as described previously,11 showed an average of 10 bacterial species in each sample, calculated at 97% similarity, and an average of six genera, calculated at 95% sequence similarity. This indicates that SSIs are typically multispecies infections, with an average of six genera in each wound and an average of 10 bacterial species.

Discussion

We hypothesised that either a single major culturable pathogen, such as S. aureus, would be associated with all such wounds or that no single pathogen is associated with them. The latter hypothesis suggests that a mixed-species biolm community, as opposed
to a lone pathogen, causes the chronic infection observed in chronic SSIs. The results indicate that a diverse population of bacteria was usually present in these samples. Most of the bacteria identied were anaerobic. These anaerobic rods may not respond to typical therapies, which are targeted at facultative aerobic cocci and are usually based on aerobic culture-based diagnostics. Our results support other studies that used molecular techniques and consistently found Bacteroides in chronic wounds.11,14-16 Corynebacterium spp. occurred in 11 of the samples and was the predominant genus in seven. Similar results were obtained in an earlier study evaluating bacterial diversity in DFUs using the same broad survey approach as here.9 The most ubiquitous genus identied was a previously uncharacterised species of Corynebacterium, which was found in 30 out of the 40 ulcers. Bacteroides and Peptoniphilus were also highly prevalent and were present in 25 and 25 samples respectively.9 Corynebacterium is an underappreciated pathogen. Other studies have associated it with diabetic foot osteomyelitis, as reviewed previously.17 In the study noted in this review, traditional culture methods identied the fastidious (hard to grow in the laboratory) Corynebacterium, which is commonly considered a contaminant, as a pathogen in DFU-associated

osteomyelitis. However, the easily cultured S. aureus is commonly considered the primary pathogen in ulcers in general. We suggest that culture techniques overestimate the importance of organisms that are easily cultured and can underestimate that of fastidious organisms, such as anaerobes and Corynebacterium, and certainly this new Bacteroidales.18 Indeed, we have been unable to culture and isolate this new Bacteroidales in the laboratory using the most advanced anaerobe culture methods. Anaerobes such as this newly discovered Bacteroidales are beginning to be recognised as major populations in chronic wound biolms.12,19-23 The importance of anaerobes such as Peptostreptococcus, Prevotella, Finegoldia and Peptoniphilus has been reported.19,24-28 This is supported by the present study, which found that these genera represent a signicant portion of SSI microbiota. Even though wounds are typically exposed to air,25 anaerobes may be the most prevalent physiological type in a given wound or an individual wound type. Bowler et al.25 evaluated infected VLUs using cultural isolation techniques that included special considerations for the propagation of anaerobes. They reported that anaerobes represented 49% of the total microbial composition. Dowd et al.,9 using a pyrosequencing approach, reported that 30% of the DNA sequences from pooled chronic DFUs were anaerobes and later conrmed that, in individual DFUs, anaerobes were also a predominant part of the biolm ecology.9 Laboratory studies have shown29,30 that obligate anaerobes may cope with the toxic effects of oxygen by interacting with aerobic or facultative bacteria populations in a symbiotic manner as part of a process known as coaggregation. Aerobic species may consume oxygen and create localised niches, allowing the obligate anaerobes to gain an advantage when in close proximity to their oxygen-reducing neighbours. Rasmussen et al. have also shown that oxygen only penetrates microns into the surface of biolms, suggesting that internal regions of the bacterial communities may support only anaerobes and facultative anaerobes.31 It has been proposed that sequencing the 16S gene of clinical, laboratory-cultivated bacteria, as is the case with bTEFAP, has advantages over traditional biochemical identication methods.32 We propose that not only is bTEFAP more effective in identifying cultivated microbes, but such assays can also be universal methods of pathogen diagnosis. Indeed, our ndings highlight the shortcoming of relying on culture methods to identify important bacterial populations within clinical samples. Infections that are predominantly caused by bacteria in the planktonic phenotype tend to be acute, with a signicant host response that is characterised
by the classic signs and symptoms including erythema, pain, swelling and heat. The hallmark of acute infections is that they are susceptible to antibiotics and resolve in 1014 days.33 Planktonic-phenotype bacteria explain much of this behaviour. Planktonic and motile bacteria upregulate virulence factors, bacterial proteases and other secreted agents to lyse tissues, on which it then feeds.34 The perceived pattern of acute planktonic infection is one of predation: if the host does not adequately respond or there is no outside intervention, the host dies. Many SSIs, however, occur after discharge and show a slow, undulating course that is considerably different to an acute infection. Often the entire incision will dehisce, but there is no degradation of tissue surrounding the wound. The damage is most often conned to the surface of the surgical incision. Biolms are more successful on surfaces, especially interfacing surfaces.35 Also, even though culture methods demonstrate at least some of the bacterial species present in biolms, antibiotics are unsuccessful in eradicating most of these infections.36 The presence of a biolm on the surface of the surgical wound may explain why we stand by helplessly while our planktonic tools and strategies fail to prevent the wound from dehiscing. Biolm management of SSIs is based on multiple concurrent strategies that specically target biolm behaviour.7 This includes opening any tunnelling and undermining by removing sutures or opening skin to expose the surface-associated bacteria. This robs the biolm of a second surface to organise around, and creates access for other strategies. The biolm can be deprived of its nutritional source by immunosuppressants, but this blocks host-healing responses and should be considered a last resort in SSIs. Topical negative pressure may be substituted as it draws off exudate, potentially limiting the nutrient supply. Frequent debridement of the wound surface forces the biolm to constantly reconstitute itself and makes it more susceptible to antibiotics and selective biocides.36,37 Using antibiotics at high doses (times above the minimum inhibitory concentration [MIC]) for 68 weeks38 (improves biolm suppression.39 As Fux39 explicitly indicated, antibiotics alone will rarely be successful against biolms and should only be used with other strategies. Selective biocides such as silver or cadexomer iodine will suppress biolm phenotype bacteria up to one half log, but do not harm host healing responses.40,41 Biolms are best managed through physical disruption. This principle has been proven in our baths, toilets and on our teeth. It has also been demonstrated in packaging, processing and pool maintenance. By frequently disrupting the biolm

with brushes or by other physical means, the colony is degraded. Then, as the colony tries to reconstitute itself, treating agents such as biocides, antibiotics and even quorum-sensing inhibitors becomes more effective. This principle can be exploited in several ways. Consider a quick opening of the involved area of the wound, removing any dead and devitalised tissue frequently, and then physically or enzymatically managing the wound surface to suppress the re-accumulation of biolm at weekly intervals. Such physical disruption of the biolm is rarely sufcient in itself. By adding other simultaneous strategies, such as selective biocides, antibiolm agents and antibiotics, it is often possible to suppress the biolm accumulation. When the biolm is suppressed, host healing processes like angiogenesis, extracellular matrix formation and wound contraction become much more effective. It has been demonstrated that, by targeting biolms, a higher percentage of chronic wounds heal, demonstrating that biolm is an important barrier to healing.7 Early intervention with aggressive, multiple, concurrent strategies that target surface-associated bacteria on the SSI may therefore improve outcomes. This study is the rst to use next-generation pathogen-detection methods to evaluate the diversity of bacteria in SSIs. Furthermore, it involved a broader
12 Dowd, S.E., Sun,Y., Secor, P.R. et al. Survey of bacterial diversity in chronic wounds using Pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC Microbiol 2008; 6: 1, 43. 13 Wolcott, R.D., Dowd, S.E. A rapid molecular method for characterising bacterial bioburden in chronic wounds. J Wound Care 2008; 27: 17, 513-516. 14 Hill, K.E., Davies, C.E., Wilson, M.J. et al. Molecular analysis of the microora in chronic venous leg ulceration. J Med Microbiol 2003; 52: 4, 365-369. 15 Redkar, R., Kalns, J., Butler, W. et al. Identication of bacteria from a non-healing diabetic foot wound by 16 S rDNA sequencing. Mol Cell Probes 2000; 14: 3, 163-169. 16 Lewis, R.P., Sutter,V.L., Finegold, S.M. Bone infections involving anaerobic bacteria. Medicine (Baltimore) 1978; 57: 4, 279-305. 17 Hartemann-Heurtier, A., Senneville, E. Diabetic foot osteomyelitis. Diabetes Metab 2008. 18 Cartwright, C.P., Stock, F., Gill, V.J. Improved enrichment broth for cultivation of fastidious organisms. J Clin Microbiol 1994; 32: 7, 1825-1826. 19 Bowler, P.G., Davies, B.J. The microbiology of infected and noninfected leg ulcers. Int J Dermatol 1999; 38: 8, 573-578. 20 Brook, I., Frazier, E.H. Aerobic and anaerobic microbiology of chronic venous ulcers. Int J Dermatol 1998; 37: 6, 426-428. 21 Brook, I. Role of encapsulated anaerobic bacteria in synergistic infections. Crit Rev Microbiol 1987;14: 3,171-193. 22 Mayrand, D., McBride, B.C. Exological relationships of bacteria involved in a simple, mixed anaerobic infection. Infect Immun 1980; 27: 1, 44-50. 23 Urbancic-Rovan,V., Gubina, M. Bacteria in supercial diabetic foot ulcers. Diabet Med 2000;17: 11, 814-815. 24 Trengove, N.J., Stacey, M.C., McGechie, D.F. et al. Qualitative bacteriology and leg ulcer healing. J Wound Care 1996; 5: 6, 277-80. 25 Bowler, P.G., Davies, B.J., Jones, S.A. Microbial involvement in chronic wound malodour. J Wound Care 1999; 8: 5, 216-218. 26 Hansson, C., Hoborn, J., Moller, A. et al. The microbial ora in venous leg ulcers without clinical signs of infection: repeated culture using a validated standardised microbiological technique. Acta Derm Venereol 1995; 75: 1, 24-30. 27 Kontiainen, S., Rinne, E. Bacteria in ulcera crurum. Acta Derm Venereol 1988; 68: 3, 240-244. 28 Howell-Jones, R.S., Wilson, M. J., Hill, K.E. et al. A review of the microbiology, antibiotic usage and

range of patients than did most other studies that used molecular methods on these wounds. Still, only 23 samples were analysed from a single geographical region. Future work should look at a larger population from different geographical regions to gain a better understanding of the microbial populations associated with such infections.

Conclusion

We have used advanced, next-generation, molecular methods to evaluate the microbial ecology of SSIs. Previous literature, which has relied on outdated laboratory culture techniques, has stated that Grampositive cocci, such as S. aureus, are dominant. In contrast, our results suggest that anaerobic rodshaped bacteria predominate in biolms. The inability to grow such bacteria using standard culture-based methods explains the historical inaccuracy of this information. Newer diagnostic methods, such as bTEFAP, may enable us to better target therapies to the pathogens in chronic wounds. Finally, we have identied two previously uncharacterised Bacteroidales as ubiquitous and primary populations in most SSIs. Work is already underway to isolate and characterise these unknown Bacteroidales and dene therapeutics for their control. This study may thus represent the initial description of a novel anaerobic pathogen associated with SSIs.
resistance in chronic skin wounds. J Antimicrob Chemother 2005; 55: 2, 143-149. 29 Bradshaw, D.J., Marsh, P.D., Watson, G.K. et al. Role of Fusobacterium nucleatum and coaggregation in anaerobe survival in planktonic and biolm oral microbial communities during aeration. Infect Immun 1998; 66: 10, 4729-4732. 30 Bradshaw, D.J., Marsh, P.D., Allison, C. et al. Effect of oxygen, inoculum composition and ow rate on development of mixedculture oral biolms. Microbiology 1996; 142: 3, 623-629. 31 Rasmussen, K., Lewandowski, Z. Microelectrode measurements of local mass transport rates in heterogeneous biolms. Biotechnol Bioeng 1998; 59: 3, 302-309. 32 Clarridge, J.E. Impact of 16S rRNA gene sequence analysis for identication of bacteria on clinical microbiology and infectious diseases. Clin Microbiol Rev 2004; 17: 4, 840-862. 33 Leibovitz, E. Acute otitis media in pediatric medicine: current issues in epidemiology, diagnosis, and management. Paediatr Drugs 2003; 5: S1, 1-12. 34 Overhage, J., Bains, M., Brazas, M.D. et al. Swarming of Pseudomonas aeruginosa is a complex adaptation leading to increased production of virulence factors and antibiotic resistance. J Bacteriol 2008; 190: 8, 2671-2679. 35 Otto, M. Staphylococcal biolms. Curr Top Microbiol Immunol 2008; 322, 207-228. 36 Stewart, P.S., Rayner, J., Roe, F. et al. Biolm penetration and disinfection efcacy of alkaline hypochlorite and chlorosulfamates. J Appl Microbiol 2001; 91: 3, 525-532. 37 Stewart, P.S. Mechanisms of antibiotic resistance in bacterial biolms. Int J Med Microbiol 2002; 292: 2, 107-113. 38 Sandoe, J.A., Kerr, K.G., Reynolds, G.W., Jain, S. Staphyloccus capitis endocarditis: two cases and review of the literature. Heart 1999; 82: 3, e1. 39 Fux, C.A., Costerton, J.W., Stewart, P.S. et al. Survival strategies of infectious biolms. Trends Microbiol 2005; 13: 1, 34-40. 40 Leaper, D. J., Durani, P. Topical antimicrobial therapy of chronic wounds healing by secondary intention using iodine products. Int Wound J 2008; 5: 2, 361-368. 41 Wiegand, C., Heinzem T., Hipler, U. C. Comparative in vitro study on cytotoxicity, antimicrobial activity, and binding capacity for pathophysiological factors in chronic wounds of alginate and silver-containing alginate. Wound Repair Regen 2009; 17: 4, 511-521.