Characteristics of microbial drug resistance and its correlates in chronic diabetic foot ulcer infections

Transcription

1 Journal of Medical Microbiology (2014), 63, DOI /jmm Characteristics of microbial drug resistance and its correlates in chronic diabetic foot ulcer infections Thokur S. Murali, 1 Shettigar Kavitha, 1 Jain Spoorthi, 1 Deepika V. Bhat, 1 Alevoor S. Bharath Prasad, 1 Zee Upton, 2 Lingadakai Ramachandra, 3 Raviraj V. Acharya 3 and Kapaettu Satyamoorthy 1 Correspondence Kapaettu Satyamoorthy ksatyamoorthy@manipal.edu 1 School of Life Sciences, Manipal University, Manipal, India 2 Tissue Repair and Regeneration Program, Institute of Health and Biomedical Innovation, Queensland University of Technology, Kelvin Grove, Queensland, Australia 3 Kasturba Medical College, Manipal, India Received 19 March 2014 Accepted 17 July 2014 While virulence factors and the biofilm-forming capabilities of microbes are the key regulators of the wound healing process, the host immune response may also contribute in the events following wound closure or exacerbation of non-closure. We examined samples from diabetic and non-diabetic foot ulcers/wounds for microbial association and tested the microbes for their antibiotic susceptibility and ability to produce biofilms. A total of 1074 bacterial strains were obtained with staphylococci, Pseudomonas, Citrobacter and enterococci as major colonizers in diabetic samples. Though non-diabetic samples had a similar assemblage, the frequency of occurrence of different groups of bacteria was different. Gram-negative bacteria were found to be more prevalent in the diabetic wound environment while Gram-positive bacteria were predominant in non-diabetic ulcers. A higher frequency of monomicrobial infection was observed in samples from non-diabetic individuals when compared to samples from diabetic patients. The prevalence of different groups of bacteria varied when the samples were stratified according to age and sex of the individuals. Several multidrug-resistant strains were observed among the samples tested and most of these strains produced moderate to high levels of biofilms. The weakened immune response in diabetic individuals and synergism among pathogenic micro-organisms may be the critical factors that determine the delicate balance of the wound healing process. INTRODUCTION One of the complications associated with diabetes is chronic foot ulcers with severe infection. Even though studies on microbial profile over time in chronic wounds are lacking, reports suggest that the microbial community in chronic wounds follows a distinct succession of associated species while few other species are consistently present (Howell- Jones et al., 2005). The diversity of microbes in infected diabetic foot ulcers may be influenced by several factors including wound depth, tissue perfusion, altered host humoral immunity (Martin et al., 2010) and geographical location. Hence, proper wound management would include not only a thorough understanding of the underlying Abbreviations: MDRO, multidrug-resistant organism; MRCONS, meticillin-resistant coagulase-negative staphylococci; MRSA, meticillin-resistant Staphylococcus aureus; MSCONS, meticillin-sensitive coagulasenegative staphylococci; MSSA, meticillin-sensitive S. aureus; OD, optical density. The GenBank/EMBL/DDBJ accession numbers for the16s 23S rrna sequences of representative strains obtained in the study are KF KF aetiology, but also knowledge of the actual prevalence of microbes and their response to antimicrobials during the course of wound management. The range of microbes in infected wounds and a number of intricate host factors that induce release of an array of pro-inflammatory effectors may play a role in influencing wound closure and healing (Dinh et al., 2012; Martin et al., 2010). In severe cases, this may even lead to amputation of limbs; hence appropriate antibacterial therapy using broad-spectrum antibiotics is often used. However, indiscriminate use of common antibiotics may eventually lead to the emergence of strains resistant to many antibiotics (Davies & Davies, 2010). This is due to the fact that the bacterial species can acquire resistance extrinsically, such as against macrolide antibiotics, by horizontal gene transfer of plasmids (Palmer et al., 2010), or by intrinsic mechanisms such as mutations and amplifications (Andersson & Hughes, 2010; Raghunath, 2008). Multidrug-resistant organisms (MDROs) are common in patients with diabetic foot ulcers (Gadepalli et al., 2006; Ramakant et al., 2011). MDROs are resistant to several G 2014 The Authors Printed in Great Britain 1377

2 T. S. Murali and others antibiotics and hence require treatment with extendedspectrum antibiotics for long durations. Infection with MDROs can cause prolonged stays in hospital, leading to higher treatment costs and thereby concomitantly increasing the chances of acquiring other nosocomial infections, leading to a risk of increased morbidity and mortality (Wang et al., 2010). Although this is common knowledge, there are very few reports on MDROs associated with diabetic foot ulcers in India. There are several factors that contribute to the emergence and survival of MDROs in diabetic foot ulcers. Chief among them are extracellular polymeric matrices (biofilms) produced by bacteria that may contribute to multidrug resistance in bacteria. Biofilms may affect effective penetration of antibiotics into the wound bed or may render leukocytes ineffective by developing antiphagocytic properties inside the biofilm matrix (Brady et al., 2008; Carson et al., 2010). The microcolonies formed in the biofilms can facilitate intercellular communication, horizontal gene transfer and altered gene expression. They may also play a vital role in quorum sensing (Srivastava et al., 2011) leading to an altered phenotype of microbes which in turn may influence wound healing. In view of this, and the availability of diverse classes of antibiotics for therapy, it becomes important that therapeutic strategies defined by specific characterization of microbes associated with the wound are rationally implemented to facilitate effective treatment of wound infection. Hence, as a prelude to metagenomic studies, the present study was undertaken to characterize common bacterial microbes associated with diabetic foot ulcer infections in a large group of clinical samples collected from patients in semi-urban and rural settings. METHODS Study design. A total of 1019 samples were collected from 357 diabetic patients with foot ulcers (284 males and 73 females) and 224 samples were collected from 104 non-diabetic individuals with nondiabetic foot ulcers (91 males and 13 females) admitted to Kasturba Medical College, Manipal between December 2010 and August Prior written informed consent was obtained from all the individuals as approved by the Ethical Committee of Manipal University. Multiple samples were obtained from patients wherever possible. Age, sex, type of diabetes, duration since first diagnosis, co-morbidities, glycaemic control during treatment, size of ulcer and details on antibiotic therapy were collected from hospital records for each patient. In the case of diabetic individuals, patients with wounds which were not aetiologically related to their diabetic state or its complications, including varicose ulcers, vasculitis, or with specific infections or neoplasms, were not included in the study. Isolation of microbes. Pus swabs, exudates and debrided tissue samples were collected from each patient following debridement of superficial exudates from the diabetic foot ulcer region. The swabs were immediately processed for culturing of aerobic bacteria. Tissue samples were also stained using Gram staining procedures. A series of standard biochemical and microbiological techniques were followed to isolate and identify bacterial strains for each sample obtained in the study (Forbes et al., 2007). Molecular identification of representative strains was carried out by either 16S rdna or 23S rdna sequencing and the sequences were deposited in the National Center for Biotechnology Information (NCBI) database. The swabs were cultured on blood agar and MacConkey agar (HiMedia Laboratories) for aerobic culture. Inoculated media were incubated at 37 uc for h for bacterial cultures. Tissues were gently homogenized before inoculation. Antimicrobial sensitivity patterns. Assessment of the antimicrobial sensitivity of the bacterial isolates was performed using the disc diffusion method. Antimicrobial discs [amikacin (30 mg), ampicillin (10 mg), aztreonam (30 mg), cefazolin (30 mg), cefepime (30 mg), cefotaxime (30 mg), cefoxitin (30 mg), ceftazidime (30 mg), cefuroxime (30 mg), ciprofloxacin (5 mg), chloramphenicol (30 mg), co-amoxiclav (30 mg), co-ticarclav (75 mg ticarcillin, 10 mg clavulanic acid), erythromycin (15 mg), gentamicin (10 mg), high level gentamicin (120 mg), netillin (30 mg), piperacillin (100 mg), rifampicin (5 mg), teicoplanin (30 mg), tetracycline (30 mg), tobramycin (10 mg)] were obtained from HiMedia Laboratories and results were interpreted as recommended by the Clinical and Laboratory Standards Institute guidelines (CLSI, 2011). For MIC studies, amikacin ( mg), ampicillin ( mg), chloramphenicol ( mg), cefazolin ( mg), cefotaxime ( mg), ceftazidime ( mg), gentamicin ( mg), linezolid ( mg), piperacillin ( mg), tobramycin ( mg) and vancomycin ( mg) HiComb MIC strips were used (HiMedia Laboratories). The following reference strains were used in the study for quality control Escherichia coli (ATCC 25922), Staphylococcus aureus (ATCC 29213), Enterococcus faecalis (ATCC 29212) and Pseudomonas aeruginosa (ATCC 27853). Detection of biofilm formation. Biofilm formation was analysed using the tissue culture plate method with an extended incubation for a period of 24 h (Mathur et al., 2006). Isolates from fresh agar were inoculated into trypticase soy broth with 1 % glucose and incubated at 37 uc for 18 h. The overnight-grown culture was diluted 1/100 with fresh medium and 0.3 ml was added to individual wells of sterile, polystyrene, 96-well, flat bottom tissue culture plates. The tissue culture plates were incubated for 24 h at 37 uc. The wells were washed four times with 0.2 ml sterile distilled water after removing the contents of each well. The biofilms were stained with 0.1 % crystal violet and solubilized using 30 % glacial acetic acid. Optical density of the stained biofilm was determined with an ELISA autoreader (Infinite M200; Tecan) at a wavelength of 570 nm. The experiments were performed in triplicate. Wells with only the broth served as controls. Statistical analysis. The quantitative variables were expressed as means±sd while qualitative variables were presented as frequencies. A x 2 test, Student s t-test and Mann Whitney test were carried out for data analysis. The statistical analyses were performed using SPSS statistical software (version 16.0). RESULTS Isolation of bacteria A total of 1019 samples from 357 diabetic patients (284 males and 73 females) were screened for the presence of microbes associated with foot ulcers. Males were predominant in both the diabetic and non-diabetic groups, contributing % of the subjects included in the study. The median age of the subjects was 58±10.36 years (Table 1). A total of 940 bacterial isolates consisting of 587 Gramnegative bacilli and 353 Gram-positive cocci were obtained. Of the samples studied in which at least one type of bacteria was isolated, 58 % of the isolates were polymicrobial in 1378 Journal of Medical Microbiology 63

3 Characterization of diabetic foot ulcer infections Table 1. Characteristics of diabetic and non-diabetic individuals included in the study Non-diabetic Number of individuals Males 284 (80 %) 91 (88 %) Females 73 (20 %) 13 (12 %) Median age 58.00± ±16.16 Age, Age Age 61 and above Strains obtained Gram-positive strains 353 (38 %) 75 (56 %) Gram-negative strains 587 (62 %) 59 (44 %) Monomicrobial infections 42 % 67 % Polymicrobial infections 58 % 33 % Random glucose (mg dl 21 ) Glucose postprandial levels (mg dl 21 ) Glucose fasting levels (mg dl 21 ) Hb A 1c (%) Hb (g dl 21 ) Neutrophils (%) Number of samples nature. Gram-positive organisms alone were found in 25 % of the samples while Gram-negative organisms alone were found in 44 % of the samples. Gram-positive and -negative organisms coexisted in 31 % of the samples. Staphylococci and Pseudomonas accounted for 26 and 22 % of all isolates obtained from diabetic foot ulcers, respectively (Fig. 1). Other species including Enterobacter, streptococci and E. coli were isolated at a frequency of less than 5 %. In the case of the control samples studied, 134 isolates were obtained from a total of 224 samples collected from 104 individuals (91 males and 13 females). Of the control samples studied, 67 % showed monomicrobial infections, while 33 % showed polymicrobial infections. A x 2 test showed that there was a significant difference between samples from diabetic and non-diabetic individuals with reference to monomicrobial and polymicrobial infections (P,0.001; x 2 value ). We found that although the mean neutrophil levels were not significantly different between the two groups (P ; t-test), there was a significant difference in the median values in these two groups (P ; Mann Whitney test). While staphylococci and Pseudomonas dominated the assemblage in both the control and diabetic samples, the frequency of occurrence was considerably higher for staphylococci in control samples (Fig. 1). There was a significant difference in the frequency of occurrence in samples from diabetic versus non-diabetic individuals for Acinetobacter (P50.002), Proteus (P50.01) and enterococci (P50.01). When the samples from diabetic individuals were stratified according to age, we found that the staphylococci and Pseudomonas were predominant in samples from individuals of years. Staphylococci were consistently present in samples across all age groups (Fig. 2). Similarly, we found more isolates belonging to the genus Citrobacter in samples from diabetic females when compared to samples from males (P50.007) (Fig. 3). We found preferential coexistence of certain bacterial species in samples isolated from diabetic foot ulcers. For example, Frequency (%) Acinetobacter Citrobacter Enterococci GNB Klebsiella Staphylococci Proteus Bacterium Pseudomonas Others Fig. 1. Occurrence of different bacteria in diabetic foot ulcer and control samples. Microbes identified from diabetic and nondiabetic patients samples were analysed as described in the Methods and represented as prevalence of each type of bacteria. Stippled bars, diabetic patients; black bars, non-diabetic patients; GNB, Gram-negative bacilli. Frequency (%) Acinetobacter Citrobacter Enterococci GNB Bacterium Klebsiella Staphylococci Pseudomonas Fig. 2. Occurrence of bacteria in diabetic foot ulcers from individuals across age groups. Samples belonging to diabetic patients were classified according to patient age and analysed as described in the Methods for various types of microbes. Groups of bacteria identified are indicated as frequency of their occurrence (in percentage). White bars, age years; stippled bars, age years; black bars, age 61 years and above; GNB, Gramnegative bacilli

4 T. S. Murali and others Frequency (%) Acinetobacter Citrobacter Enterococci GNB Bacterium Klebsiella Staphylococci Pseudomonas Fig. 3. Occurrence of different bacteria in diabetic foot ulcer samples from male and female individuals. Samples belonging to diabetic patients were classified according to sex and analysed as described in the Methods for various types of microbes. Groups of bacteria identified are indicated as frequency of their occurrence (in percentage). Stippled bars, males; black bars, females; GNB, Gram-negative bacilli. we found most of the bacterial species coexisted with Pseudomonas aeruginosa (Table 2). However, this could be explained by the fact that the frequency of occurrence of Pseudomonas aeruginosa was higher than that of other bacterial species. We also found that Proteus vulgaris coassociated with enterococci in more samples, and similarly, Citrobacter coexisted with enterococci in several samples. Interestingly, we found that none of the isolates belonging to Citrobacter koseri were found alone in a wound sample, while Proteus mirabilis preferred to coexist with other bacterial strains. Pseudomonas aeruginosa, meticillin-resistant coagulase-negative staphylococci (MRCONS) and meticillinresistant Staphylococcus aureus (MRSA) occurred alone more frequently than other bacteria in the samples studied. In non-diabetic wounds, none of the isolates belonging to Klebsiella pneumoniae and Proteus mirabilis occurred alone (Table 3). Antimicrobial resistance profile Antimicrobial resistance profiles were studied for the isolated strains except for unidentified species. The CLSI guidelines were followed for testing and interpretation of drug susceptibility and multiple antibiotics were used for the antimicrobial susceptibility assay. The results are presented in Table 4. The results were confirmed by studying MIC values using HiComb strips (data not shown). Among the antimicrobials tested, members of the Enterobacteriaceae family showed high resistance against b-lactams and least resistance against aminoglycosides (Table 4). More than 90 % of strains belonging to Acinetobacter spp. were resistant to all the b-lactams tested. The majority of Acinetobacter strains had developed resistance to most of the common antibiotics and the most effective drug was found to be netillin (57 and 25 % resistance in diabetic and non-diabetic subjects, respectively). Drug resistance to most of the antibiotics, including cephalosporins, was greater than 50 %. In the case of Pseudomonas aeruginosa, in diabetic patients, piperacillin was the most effective drug with only 15 % of strains showing resistance while amikacin was effective Table 2. Percentage of isolates of bacteria found associated with other bacterial isolates in diabetic wound ulcer samples Only those bacteria that were isolated from more than 20 samples are shown. For each species, only percentage values above 10 % are shown. ACI, Acinetobacter spp.; CIB, Citrobacter spp.; CIK, Citrobacter koseri; ENB, Enterobacter spp.; ENC, Enterococci; GNB, Gram-negative bacilli; KLP, Klebsiella pneumonia; MRC, meticillin-resistant coagulase-negative staphylococci; MRS, meticillin-resistant S. aureus; MSS, meticillin-sensitive S. aureus; PMI, Proteus mirabilis; PSA, Pseudomonas aeruginosa; PVU, Proteus vulgaris. Bacterial species Coexisting with ACI CIB CIK ENB GNB KLP PMI PVU PSA ENC MSS MRC MRS % Alone % Others ACI CIB CIK ENB GNB KLP PMI PVU PSA ENC MSS MRC MRS Journal of Medical Microbiology 63

5 Characterization of diabetic foot ulcer infections Table 3. Percentage of isolates of bacteria found associated with other bacterial isolates in non-diabetic wound ulcer samples ACI, Acinetobacter spp.; EC, Escherichia coli; ENB, Enterobacter spp.; ENC, Enterococci; GNB, Gram-negative bacilli; KLP, Klebsiella pneumonia; MRC, meticillin-resistant coagulase-negative staphylococci; MRS, meticillin-resistant S. aureus; MSC, meticillin-sensitive coagulase-negative staphylococci; MSS, meticillin-sensitive S. aureus; PMI, Proteus mirabilis; PSA, Pseudomonas aeruginosa. Bacterial species Coexisting with ACI ENB GNB EC KLP PMI PSA ENC MSS MSC MRC MRS % Alone % Others ACI ENB GNB EC KLP PMI PSA ENC MSS MSC MRC MRS against 70 % of Pseudomonas aeruginosa strains tested. Fourteen of the Pseudomonas aeruginosa strains showed 100 % sensitivity to all the antimicrobials tested while nine strains were pan-resistant. Further, of the Pseudomonas aeruginosa isolates tested for their antibiotic sensitivity, the highest resistance was shown to cefepime, and 81 % of these cefepime-resistant strains were susceptible to piperacillin. Among the staphylococci tested using the disc diffusion assay, 67 % of the strains showed resistance to cefoxitin in both diabetic and non-diabetic wounds (Table 4). More than 65 % of the strains were resistant to all of the three b- lactam drugs tested (co-amoxiclav, ampicillin and cefoxitin). The incidence of resistance to different antibiotics in staphylococci varied from 17 % (chloramphenicol) to more than 90 % (ampicillin). In diabetic patients, only two strains were found to be sensitive to all the antibiotics tested and these belonged to the meticillin-sensitive S. aureus (MSSA) and meticillin-sensitive coagulase-negative staphylococci (MSCONS) groups. Interestingly, MRCONS isolates (10 %) were more frequently isolated than MRSA (7 %) in samples from diabetic wounds, while more MRSA strains were observed in samples from non-diabetic wounds (17 % MRSA versus 11 % MRCONS). In our study, Enterobacteriaceae and staphylococci showed high resistance to ampicillin (around 90 %), but enterococci were found to be more sensitive to the same antibiotic (78 % of strains were sensitive). In the case of enterococci from diabetic samples, more than 87 % of the strains tested were resistant to tetracycline and erythromycin. However, 37 % of enterococci were found to show high levels of gentamicin resistance. We found a significant difference in microbial resistance to ampicillin for enterococci isolated from diabetic and non-diabetic samples (Table 4). Detection of biofilm formation A total of 355 strains of micro-organisms obtained from both diabetic and non-diabetic patients were tested for their ability to form biofilms (Tables 5 and 6, respectively). Based on the tissue culture plate assay, the strains were classified as high [optical density (OD).0.40], moderate (OD between ) or low (OD,0.20) biofilm-forming organisms. It was shown that among the total 355 isolates, 159 strains had the ability to form high to moderate levels of biofilm. Out of these, 94 strains were classified as high biofilmproducing organisms and the remaining 65 strains were grouped as moderate biofilm producers. The remaining 196 strains were considered as non/weak biofilm producers. Interestingly, we found that most of the biofilm-forming strains were multidrug-resistant (resistant to more than three antibiotics). In the present study, Pseudomonas aeruginosa contributed to a higher proportion of strains in the diabetic wound beds compared with other bacteria, albeit only 8 % of the 51 strains tested had the ability to form significantly high amounts of biofilm (Table 5). Three strains belonging to Pseudomonas aeruginosa were resistant to all the antibiotics tested and among them, two strains were shown to be low biofilm producers while one strain produced high levels of biofilm. These three Pseudomonas aeruginosa strains were all isolated from diabetic wounds. Though all the 23 strains of Acinetobacter were found to be multidrug-resistant, only nine strains were found to be high biofilm producers (Table 5). In the case of bacteria isolated from non-diabetic samples, more than 50 % of staphylococci were high biofilm producers and were also multidrug-resistant (Table 6). Only in the case of E. coli was a significant difference in the biofilm production observed in isolates between diabetic and non-diabetic wounds, even though the numbers of strains tested were lower in the case of

6 1382 Journal of Medical Microbiology 63 Table 4. Antimicrobial resistance profile of bacteria isolated from diabetic and non-diabetic foot ulcer samples Values are given as percentages. ACI, Acinetobacter spp.; ENB, Enterobacteriaceae; ENC, Enterococci; PSA, Pseudomonas aeruginosa; STA, Staphylococci. Antimicrobial Nondiabetic Nondiabetic Nondiabetic Nondiabetic Nondiabetic ENB (n5206) ENB (n516) ACI (n551) ACI (n54) PSA (n5150) PSA (n58) STA (n5232) STA (n554) ENC (n546) ENC (n56) Amikacin Ampicillin * 83* Aztreonam Cefazolin Cefepime Cefotaxime Cefoxitin Ceftazidime Cefuroxime Chloramphenicol Ciprofloxacin Co-amoxiclav Co-ticarclav Erythromycin Gentamicin High level gentamicin Linezolid 2D Netillin Piperacillin Rifampicin 63 0 Teicoplanin Tetracycline Tobramycin Vancomycin 2D *P (significant difference between diabetic and non-diabetic samples). DMIC values for 41 strains tested. T. S. Murali and others

7 Characterization of diabetic foot ulcer infections Table 5. Low, moderate and high biofilm-forming strains along with the multidrug resistance profile among different groups of bacteria in diabetic patients Values are given as percentages. H-MD, High biofilm-forming MDRO strains; H-NM, high biofilm-forming non-mdro strains; L-MD, low biofilm-forming MDRO strains; L-NM, low biofilm-forming non-mdro strains; M-MD, moderate biofilm-forming MDRO strains; M-NM, moderate biofilm-forming non-mdro strains. samples (n5328) L-MD L-NM M-MD M-NM H-MD H-NM Acinetobacter spp. (23) Citrobacter spp. (47) Enterobacter spp. (27) Enterococci (50) Escherichia coli (8)* GNB (4) Klebsiella spp. (23) Staphylococci (66) Proteus spp. (27) Pseudomonas aeruginosa (51) Streptococci spp. (2) *P, (significant difference between diabetic and non-diabetic groups). non-diabetic wounds compared with diabetic wounds (Tables 5 and 6). DISCUSSION individuals tend to develop several complications such as neuropathy and peripheral arterial disease that contribute to foot ulceration (Dinh et al., 2012). In the present study, we have attempted to characterize diabetic foot ulcer infections using conventional microbiological and biochemical methods of isolation and identification of microbes. In our microbiological analysis of samples from diabetic foot ulcer and non-diabetic foot wounds, we isolated a total of 1074 bacterial isolates from 461 individuals. Several earlier studies have shown Enterobacteriaceae members to be consistently associated with diabetic foot ulcers (Citron et al., 2007; Louie et al., 1976; Zubair et al., 2011). Among Enterobacteriaceae members, we found Citrobacter and Proteus to be dominant genera, followed by Klebsiella. We found Gram-negative bacteria to be more prevalent than Gram-positive bacteria, as has been observed in other earlier studies from India (Gadepalli et al., 2006; Ramakant et al., 2011; Zubair et al., 2011). However, interestingly, we found Gram-positive bacteria to be predominant in the control samples (Table 1). We also found marginal but significant variation in colonization of bacteria in samples from different age groups of individuals and in males and females (Figs 2 and 3). However, in both diabetic and control samples, we found staphylococci to dominate the assemblage. Earlier studies have also shown S. aureus, Staphylococcus epidermidis and Pseudomonas aeruginosa as common bacterial species frequently associated with wound fluids in diabetic foot ulcers (Howell-Jones Table 6. Low, moderate and high biofilm-forming strains along with the multidrug resistance profile among different groups of bacteria in non-diabetic patients Values are given as percentages. H-MD, High biofilm-forming MDRO strains; H-NM, high biofilm-forming non-mdro strains; L-MD, low biofilm-forming MDRO strains; L-NM, low biofilm-forming non-mdro strains; M-MD, moderate biofilm-forming MDRO strains; M-NM, moderate biofilm-forming non-mdro strains. Non-diabetic samples (n527) L-MD L-NM M-MD M-NM H-MD H-NM Enterobacter spp. (3) Enterococci (3) Escherichia coli (1)* Klebsiella spp. (3) Staphylococci (15) Proteus spp. (2) *P, (significant difference between diabetic and non-diabetic groups)

8 T. S. Murali and others et al., 2005; Ramakant et al., 2011; Wang et al., 2010; Xu et al., 2007; Zubair et al., 2011). Though a variety of micro-organisms are associated with diabetic foot ulcers, some studies have suggested that involvement of a specific micro-organism may lead to delayed wound healing. Trengove et al. (1996) suggested that the presence of polymicrobes in the wound site rather than the mere occurrence of specific micro-organisms eventually leads to delays in wound healing. Indeed, the polymicrobial nature of diabetic foot infections is well documented in the literature (Dowd et al., 2008; Ramakant et al., 2011). We also found in our study that there were more cases of polymicrobial than monomicrobial infection in diabetic wound samples, while samples from non-diabetic wounds were mostly monomicrobial (Table 1). Though bacterial load may play a critical role in the wound healing process per se, the antibiotic resistance profile of microbes associated with wound fluid may also play a significant role. Even though most of the b-lactams were not effective against members of Enterobacteriaceae, Acinetobacter and staphylococci, we found that piperacillin was the most potent antibiotic against Pseudomonas aeruginosa. Our studies also showed that only 17 % of the Pseudomonas aeruginosa isolates from diabetic wounds were piperacillin resistant a similar trend that was reported by Jones (2001). Paterson et al. (2005) found piperacillin/ tazobactam and amikacin to be the active agents against Pseudomonas. Ciprofloxacin is considered to be the most effective drug for Pseudomonas aeruginosa infections (Paterson et al., 2005); however, 46 % of our strains from diabetic wounds showed resistance to this antibiotic. Though resistance to most of the b-lactams is well documented for Pseudomonas aeruginosa, resistance to fourth generation cephalosporins is particularly alarming. However, we found third generation cephalosporins (ceftazidime) to be comparatively effective with only 41 % of strains from diabetic samples and 38 % of strains from non-diabetic samples showing resistance. Gales et al. (2001) in their surveillance study also observed a similar trend in the Asia Pacific region with Pseudomonas aeruginosa strains showing higher susceptibility to ceftazidime compared to cefepime. In the present study, 85 % of strains belonging to staphylococci were MDROs. Of all the staphylococci strains, 27 % were found to be MRSA and 37 % to be MRCONS. Though ciprofloxacin and vancomycin were considered to be highly effective against MRSA, many of these strains have now developed resistance against both these antimicrobials and incidences of vancomycin-resistant enterococci and MRSA are increasing worldwide (Jones, 2001). In a similar study, Gadepalli et al. (2006) found enterococci with high levels of resistance to erythromycin, tetracycline, and ciprofloxacin and with low level resistance to high levels of aminoglycosides. Though enterococci are commonly referred to as commensals, in diabetic individuals they can act as opportunistic pathogens (Citron et al., 2007). Several studies have shown that chronic wounds harbour biofilm-forming microbes (James et al., 2008). It is well known that microbes with biofilm-producing capability are more resistant to antibiotics due to the physiology of biofilms. The presence of multispecies communities in biofilms may play a critical role in the wound healing process; hence we studied the in vitro biofilm-forming capacity of microbes associated with diabetic foot ulcers. We observed that out of the 355 strains tested, 159 strains showed moderate (65) to high (94) level biofilm-producing phenotype. We also found most of the high and moderate biofilm-forming organisms to be MDROs (Tables 5 and 6). Dowd et al. (2008) suggested that bacterial species incapable of causing chronic infections individually can act synergistically with other co-occurring bacterial species to form biofilm communities that may substantially alter the chronicity of the wound. Further studies are warranted to clearly elucidate the relationship between microbial species, biofilm-forming capabilities and their antibiotic sensitivities. It is interesting to note that the number of microbial isolates obtained from non-diabetic wounds is far lower than from diabetic foot ulcers. This suggests an intimate relationship between host factors and microbes that infect diabetic wounds. However, the mere presence of microbes in diabetic foot ulcers may not imply a definitive role in delayed wound healing (Bowler et al., 2001). We have previously reported that neutrophils (responsible for innate immunity) from diabetic subjects lose their ability to form neutrophil extracellular traps that are responsible for eliminating microbes (Joshi et al., 2013). Cytokines such as interleukin-6, which is elevated in diabetic patients, appear to play a role in these processes in contrast to what was observed in normal individuals (Joshi et al., 2013). In our study, we also found a wide variation in neutrophil levels in diabetic individuals compared to control samples and this may play a significant role in hindering the wound healing process. Interestingly, we found culturenegative wounds were higher in non-diabetic subjects (27 %) than in diabetic subjects (14 %). This could be the reason for elevated neutrophil levels in diabetic foot ulcer patients compared with non-diabetic subjects since in bacterial infections, neutrophil levels are known to be increased in response to infection. Clearly, we need a greater understanding of the effect of microbes on wound healing, as well as an unravelling of the critical host factors that are involved in overcoming infections in diabetic wounds. ACKNOWLEDGEMENTS The study was supported by TIFAC-CORE in Pharmacogenomics, the Indo Australia Biotechnology Fund, the Department of Biotechnology, the Government of India and Manipal University, Manipal, India. The authors thank Mr Vasudeva Guddattu for help in statistical analysis and Mr Ganesh and Ms Divya for help rendered in sample processing Journal of Medical Microbiology 63

9 Characterization of diabetic foot ulcer infections REFERENCES Andersson, D. I. & Hughes, D. (2010). Antibiotic resistance and its cost: is it possible to reverse resistance? Nat Rev Microbiol 8, Bowler, P. G., Duerden, B. I. & Armstrong, D. G. (2001). Wound microbiology and associated approaches to wound management. Clin Microbiol Rev 14, Brady, R. A., Leid, J. G., Calhoun, J. H., Costerton, J. W. & Shirtliff, M. E. (2008). Osteomyelitis and the role of biofilms in chronic infection. FEMS Immunol Med Microbiol 52, Carson, L., Gorman, S. P. & Gilmore, B. F. (2010). The use of lytic bacteriophages in the prevention and eradication of biofilms of Proteus mirabilis and Escherichia coli. FEMS Immunol Med Microbiol 59, Citron, D. M., Goldstein, E. J. C., Merriam, C. V., Lipsky, B. A. & Abramson, M. A. (2007). Bacteriology of moderate-to-severe diabetic foot infections and in vitro activity of antimicrobial agents. J Clin Microbiol 45, CLSI (2011). Performance Standards for Antimicrobial Susceptibility Testing; 21st Informational Supplement. Wayne, PA: Clinical and Laboratory Standards Institute. Davies, J. & Davies, D. (2010). Origins and evolution of antibiotic resistance. Microbiol Mol Biol Rev 74, Dinh, T., Tecilazich, F., Kafanas, A., Doupis, J., Gnardellis, C., Leal, E., Tellechea, A., Pradhan, L., Lyons, T. E. & other authors (2012). Mechanisms involved in the development and healing of diabetic foot ulceration. Diabetes 61, Dowd, S. E., Wolcott, R. D., Sun, Y., McKeehan, T., Smith, E. & Rhoads, D. (2008). Polymicrobial nature of chronic diabetic foot ulcer biofilm infections determined using bacterial tag encoded FLX amplicon pyrosequencing (btefap). PLoS ONE 3, e3326. Forbes, B. A., Sahm, D. F. & Weissfeld, A. S. (2007). Bailey & Scott s Diagnostic Microbiology, 12th edn. St. Louis: Elsevier Science Health Science Division. Gadepalli, R., Dhawan, B., Sreenivas, V., Kapil, A., Ammini, A. C. & Chaudhry, R. (2006). A clinico-microbiological study of diabetic foot ulcers in an Indian tertiary care hospital. Diabetes Care 29, Gales, A. C., Jones, R. N., Turnidge, J., Rennie, R. & Ramphal, R. (2001). Characterization of Pseudomonas aeruginosa isolates: occurrence rates, antimicrobial susceptibility patterns, and molecular typing in the global SENTRY Antimicrobial Surveillance Program, Clin Infect Dis 32 (Suppl 2), S146 S155. Howell-Jones, R. S., Wilson, M. J., Hill, K. E., Howard, A. J., Price, P. E. & Thomas, D. W. (2005). A review of the microbiology, antibiotic usage and resistance in chronic skin wounds. J Antimicrob Chemother 55, James, G. A., Swogger, E., Wolcott, R., Pulcini, E. L., Secor, P., Sestrich, J., Costerton, J. W. & Stewart, P. S. (2008). Biofilms in chronic wounds. Wound Repair Regen 16, Jones, R. N. (2001). Resistance patterns among nosocomial pathogens: trends over the past few years. Chest 119 (Suppl), 397S 404S. Joshi, M. B., Lad, A., Bharath Prasad, A. S., Balakrishnan, A., Ramachandra, L. & Satyamoorthy, K. (2013). High glucose modulates IL-6 mediated immune homeostasis through impeding neutrophil extracellular trap formation. FEBS Lett 587, Louie, T. J., Bartlett, J. G., Tally, F. P. & Gorbach, S. L. (1976). Aerobic and anaerobic bacteria in diabetic foot ulcers. Ann Intern Med 85, Martin, J. M., Zenilman, J. M. & Lazarus, G. S. (2010). Molecular microbiology: new dimensions for cutaneous biology and wound healing. J Invest Dermatol 130, Mathur, T., Singhal, S., Khan, S., Upadhyay, D. J., Fatma, T. & Rattan, A. (2006). Detection of biofilm formation among the clinical isolates of staphylococci: an evaluation of three different screening methods. Indian J Med Microbiol 24, Palmer, K. L., Kos, V. N. & Gilmore, M. S. (2010). Horizontal gene transfer and the genomics of enterococcal antibiotic resistance. Curr Opin Microbiol 13, Paterson, D. L., Rossi, F., Baquero, F., Hsueh, P.-R., Woods, G. L., Satishchandran, V., Snyder, T. A., Harvey, C. M., Teppler, H. & other authors (2005). In vitro susceptibilities of aerobic and facultative Gram-negative bacilli isolated from patients with intraabdominal infections worldwide: the 2003 Study for Monitoring Antimicrobial Resistance Trends (SMART). J Antimicrob Chemother 55, Raghunath, D. (2008). Emerging antibiotic resistance in bacteria with special reference to India. J Biosci 33, Ramakant, P., Verma, A. K., Misra, R., Prasad, K. N., Chand, G., Mishra, A., Agarwal, G., Agarwal, A. & Mishra, S. K. (2011). Changing microbiological profile of pathogenic bacteria in diabetic foot infections: time for a rethink on which empirical therapy to choose? Diabetologia 54, Srivastava, D., Harris, R. C. & Waters, C. M. (2011). Integration of cyclic di-gmp and quorum sensing in the control of vpst and apha in Vibrio cholerae. J Bacteriol 193, Trengove, N. J., Stacey, M. C., McGechie, D. F. & Mata, S. (1996). Qualitative bacteriology and leg ulcer healing. J Wound Care 5, Wang, S.-H., Sun, Z.-L., Guo, Y.-J., Yang, B.-Q., Yuan, Y., Wei, Q. & Ye, K.-P. (2010). Meticillin-resistant Staphylococcus aureus isolated from foot ulcers in diabetic patients in a Chinese care hospital: risk factors for infection and prevalence. J Med Microbiol 59, Xu, L., McLennan, S. V., Lo, L., Natfaji, A., Bolton, T., Liu, Y., Twigg, S. M. & Yue, D. K. (2007). Bacterial load predicts healing rate in neuropathic diabetic foot ulcers. Diabetes Care 30, Zubair, M., Malik, A. & Ahmad, J. (2011). Clinico-microbiological study and antimicrobial drug resistance profile of diabetic foot infections in North India. Foot (Edinb) 21,