These authors contributed equally to this work.
A combination of methods to evaluate biofilm production may help to determine the clinical relevance of Staphylococcus in blood cultures
Article first published online: 22 DEC 2010
© 2010 The Societies and Blackwell Publishing Asia Pty Ltd
Microbiology and Immunology
Volume 55, Issue 1, pages 28–33, January 2011
How to Cite
Iorio, N. L. P., Lopes, A. P. d. C. N., Schuenck, R. P., Barcellos, A. G., Olendzki, A. N., Lopez, G. L. and dos Santos, K. R. N. (2011), A combination of methods to evaluate biofilm production may help to determine the clinical relevance of Staphylococcus in blood cultures. Microbiology and Immunology, 55: 28–33. doi: 10.1111/j.1348-0421.2010.00288.x
- Issue published online: 22 DEC 2010
- Article first published online: 22 DEC 2010
- Accepted manuscript online: 23 NOV 2010 12:20AM EST
- Received 11 June 2010; revised 7 October 2010; accepted 21 October 2010.
- biofilm detection methods;
Staphylococcus is the most prevalent pathogen causing bacteremia and many of its isolates possess the ability to form biofilm. In this study Staphylococcus isolates from the blood of patients with bacteremia were analyzed by two biofilm detection phenotypic methods: Congo red agar (CRA) and microtiter-plate adherence (MPA) in relation to the presence of ica genes, detected by PCR. Their oxacillin susceptibility was also evaluated. Among 127 isolates evaluated, 47 were S. aureus and 80 were coagulase negative staphylococci (CNS). Seventy-four (58.3%) isolates were mecA gene positive (27.7%S. aureus and 76.3% CNS isolates). Among the 40 S. aureus isolates which were positive for the ica genes, 25 (62.5%) were positive in MPA and 27 (67.5%) in CRA, whereas both methods combined detected 34 (85%) isolates as biofilm producers. Among 12 S. epidermidis isolates carrying ica genes, 8 were positive in MPA and 5 in CRA. The combination of CRA and MPA methods provided a better prediction of the presence of ica genes in S. aureus isolates than did either method alone.
Clinical Laboratory Standards Institute
coagulase negative staphylococci
Congo red agar
polymerase chain reaction
tryptic soy broth
- 24 hr/G
plates with 1% of glucose and incubation for 24 hr at 35°C
- 72 hr/G
plates with 1% of glucose and incubation for 24 hr at 35°C
In bacterial bloodstream infection, one of the most serious of health care associated events, Staphylococcus is the most commonly isolated pathogen, being responsible for about 38% of such infections (1). Together with S. aureus, CNS (in particular S. epidermidis), are the most frequently isolated species (2). Although they do lead to longer periods of hospitalization and higher cost of treatment, whether or not CNS are a principal cause of morbidity and mortality is uncertain. In addition, the remarkable ability of both S. aureus and CNS to acquire antibiotic resistance, especially to oxacillin, limits therapy options and consequently may increase patient morbidity and mortality (3).
Staphylococcus is an organism that commonly forms biofilm and ica genes seem to be essential for the production of this exopolysaccharide (4). It has been suggested that both biofilm and ica genes are markers of clinical significance, based on the findings that isolates from infection are more able to produce biofilm than commensal isolates (5, 6). Additionally, the presence of indwelling medical devices in patients can induce bacterial biofilm formation, which results in persistent infection (7).
Some phenotypic methods such as microtiter plate adherence (8–10) and Congo red agar cultivation (6, 10, 11) have been used to evaluate biofilm production by staphylococci. However, the reported results have been discordant, even when isolates presenting ica genes have been analyzed. Moreover, there are few studies that describe a combination of methods for analyzing biofilm formation in staphylococcal isolates from bacteremia specimens. Thus, in order to clarify the relevance of staphylococci in blood cultures, using different methods we evaluated isolates, including some that were considered to be contaminants, in relation to biofilm formation. Their oxacillin susceptibility was also evaluated.
MATERIALS AND METHODS
Setting and bacterial isolates
One hundred and twenty seven staphylococcal isolates from blood cultures which were collected consecutively between September 2001 and October 2002 from 127 patients (one isolate per each patient) at the Marcílio Dias Hospital, a Brazilian tertiary institution with 500 beds, located in Rio de Janeiro city, were analyzed. Professionals of the Commission of Hospital Infection Control considered 85 of the cases of bacteremia to be clinically significant (12), these were called true bacteremia, and the others false bacteremia. Staphylococcal identification was performed according to Iorio and coworkers (13).
CRA plates were prepared adding 0.08% (w/v) Congo red (BDH Chemicals, Poole, England), 5% (w/v) sucrose (Reagen, Rio de Janeiro, Brazil) in brain-heart infusion agar (Difco Laboratories, Detroit, MI, USA) (14, 15). The plates were incubated at 35°C for 24 hr, without (CRA-24 hr) and with 1% (w/v) of glucose (CRA-24 hr/G) and then at room temperature to a total of 72 hr (CRA-72 hr/G). Slime-producing isolates form black colonies, whereas non-producing isolates form red colonies (14). Biofilm-producer S. epidermidis strain ATCC 35984 (formerly RP62A); and its mutant HAM 892, which lacks the ability to produce biofilm (16) (donated by L. Baldassarri, Instituto Superiore di Sanità, Rome, Italy) were used as positive and negative controls, respectively.
MPA was performed according to Galliani and coworkers (17), with modifications. Four wells of a sterile 96-well flat-bottomed plastic tissue culture plate with a lid (Immulon 2, Dynatech Laboratories, Chantilly, VA, USA) were filled with 100 μL of bacterial suspension (equivalent to 0.5 McFarland standard) in TSB (Difco Laboratories) supplemented with 1% (w/v) glucose. The plates were incubated aerobically for 24 hr at 35°C. The contents of each well were then removed, and each well washed four times with 100 μL of phosphate-buffered saline (pH 7.2). The adherent bacteria were stained for 1 min with 100 μL of 0.1% (w/v) safranine per well. Excess stain was rinsed off by placing the plates under running tap water. S. epidermidis ATCC 35984 and HAM 892 comprised the controls. After the plates had been inverted and air dried, the dye was resolubilized with 50 μL of a mixed solution [95% (v/v) ethanol and 5% (v/v) ether] per well and the volume completed with 50 μL of phosphate-buffered saline. The OD of each well was measured at 490 nm with an ELISA Auto Reader (model 550, Bio-Rad Microplate Reader, Bio-Rad Laboratories, Hercules, CA, USA). The comparative analyses were performed according to Stepanovic and colleagues (18), using as negative controls the wells inoculated with the HAM 892 strain. The isolates was classified as strong (OD ≥ 0.29), moderate (0.15 < OD > 0.28), weak (0.08 < OD > 0.14) and non producer of biofilm (OD ≤ 0.07).
Detection of the ica gene
Oxacillin susceptibility testing
Susceptibility to oxacillin (Oxoid, Hampshire, England) was evaluated by the disk diffusion test on Müeller-Hinton agar plates (Difco Laboratories) according to the guidelines of the Clinical Laboratory Standards Institute (21). Furthermore, all isolates were plated on Müeller-Hinton agar supplemented with 4% (w/v) NaCl containing oxacillin (Sigma Chemical, St Louis, MO, USA) at a concentration of 6 μg/mL for S. aureus (22) or 4 μg/mL for CNS isolates (23). S. aureus ATCC 29213 (oxacillin susceptible) and S. aureus ATCC 33591 (oxacillin resistant) were included as control organisms. Oxacillin-resistant isolates were confirmed by PCR detection of the mecA gene (24) through the use of S. aureus ATCC 33591 (mecA+) and S. aureus ATCC 25923 (mecA−) as controls.
All comparisons were performed using the χ2 test. Differences between means were considered significantly different when values of P < 0.05 were obtained.
As shown in Table 1, there were 47 (37%) S. aureus and 80 (63%) CNS isolated from a total of 127 blood cultures. Seventy-four (58.3%) isolates were mecA positive, comprising 13 of the 47 S. aureus (27.7%) and 61 of the 80 CNS isolates (76.3%). S. epidermidis was the CNS species with the highest rate of oxacillin resistance (86.1%). All mecA positive S. aureus and CNS isolates were detected by an agar screening test using 6 μg/mL and 4 μg/mL of oxacillin, respectively. The oxacillin disk diffusion test had 97% sensitivity and 96% specificity (two mecA gene positive isolates were sensitive and two mecA gene negative isolates were detected as resistant). In regard to oxacillin susceptibility, no statistical difference (P > 0.05) was found between isolates from true and false bacteremia.
|Staphylococcus species (no. of isolates)||Resistant isolates/Total of isolates (% of resistance)||No. (%) of isolates mecA gene positivec|
|True bacteremia||False bacteremia||Total|
|S. aureus (47)||12/46 (26.1%)||13/46 (28.3%)||0/1||0/1||12/47 (25.5%)||13/47 (27.7%)||13 (27.7%)|
|S. epidermidis (36)||14/17 (82.3%)||15/17 (88.2%)||16/19 (84.2%)||16/19 (84.2%)||30/36 (83.3%)||31/36 (86.1%)||31 (86.1%)|
|S. haemolyticus (23)||10/13 (76.9%)||10/13 (76.9%)||8/10 (80%)||8/10 (80%)||18/23 (78.3%)||18/23 (78.3%)||18 (78.3%)|
|S. hominis (14)||4/5 (80%)||3/5 (60%)||7/9 (77.8%)||6/9 (66.7%)||11/14 (78.6%)||9/14 (64.3%)||9 (64.3%)|
|Others CNS (7)||3/4 (75%)||3/4 (75%)||0/3||0/3||3/7 (42.9%)||3/7 (42.9%)||3 (42.9%)|
The analysis of MPA, CRA cultivation and ica gene detection is showed in Table 2. Among 47 S. aureus isolates, 85.1% (40 isolates) carried an ica gene. Twenty five (62.5%) of them were positive by MPA and were classified as strong (3 isolates, 12%), moderate (3, 12%) and weak (19, 76%). Regarding CRA growth, 6 (15%) S. aureus isolates were positive after 24 hr of incubation without glucose, and 11 (27.5%) were positive after 24 hr with glucose (CRA-24 hr/G), the latter including the previous isolates. Finally a total of 27 (67.5%) isolates, these also including the previous isolates, became positive after 72 hr of incubation on the medium with added glucose (CRA-72 hr/G). The increase in positivity observed with CRA-72/G was statistically significant in relation to CRA-24 hr and CRA-24 hr/G (P < 0.05). Figure 1 shows the color adopted for colony evaluation on CRA.
|Staphylococcus species||Bacteremiaa (no. of isolates)||No. of isolates positive for:|
|Microtiter-plate adherence||Congo red agarb||MPA and CRAc||MPA and/or CRAd||ica genese|
|S. aureus||T (46)||3||3||19||26||18||34||39|
|S. epidermidis||T (17)||4||0||3||4||4||7||7|
|S. haemolyticus||T (13)||0||0||4||0||0||4||ND|
|S. hominis||T (5)||0||0||0||0||0||0||ND|
|Others CNS||T (4)||0||1f||0||0||0||1f||ND|
An increased percentage of biofilm positive isolates (85%, 34 isolates) was observed when we combined the results of both phenotypic methods. All three S. aureus isolates that were strong biofilm-producers in MPA were also positive on CRA plates. However, for three moderate and 19 weak biofilm producer isolates, one and six, respectively, did not grow on CRA. On the other hand, among 27 isolates that were positive on CRA, nine were negative with MPA. Among S aureus, only six isolates carried an ica gene.
For 36 S. epidermidis isolates, only 33.3% (12 isolates) carried ica genes (Table 2). Eight (66.7%) of them were positive in MPA and were classified as strong (4 isolates; 50%) and weak (4; 50%). Regarding CRA growth, only five (41.7%) isolates were positive. A slightly increased percentage of biofilm positive isolates (75%, 9 isolates) was observed when we combined the results of both phenotypic methods. Two of four S. epidermidis isolates that were strong biofilm-producers in MPA were negative on CRA plates. This was also true of the four weak biofilm-producers isolates. Of the five isolates that were positive on CRA, one was negative with MPA. Among S. epidermidis, three isolates only carried ica genes.
Among nine S. epidermidis biofilm producer isolates, seven (77.8%) were from true bacteremia (P < 0.05). Moreover, all four S. epidermidis isolates that were simultaneously ica genes, MPA and CRA positive originated from true bacteremias (P < 0.05). Figure 2 shows the correlation between true and false bacteremias by S. epidermidis and biofilm-production testing. Although three biofilm non-producer isolates did present ica genes, this profile was not observed among the S. epidermidis isolates from true bacteremia (P < 0.05).
All S. aureus and S. epidermidis biofilm producer isolates carried ica genes while negative isolates for these genes were also negative for both phenotypic methods analyzed in this study.
For 44 non-S. epidermidis CNS isolates, no statistical difference (P > 0.05) was observed for biofilm production between isolates from true and false bacteremias.
Among 13 S. aureus isolates that were positive for mecA gene, 10 (76.9%) were also positive for the ica gene, while among 34 sensitive isolates, 29 (85.3%) presented this gene. In relation to S. epidermidis, 11 (35.5%) from 31 isolates positive for the mecA gene carried ica genes and only one (20%) of 5 sensitive isolates was ica genes positive. Thus, there was no significant correlation between oxacillin resistance and the presence of ica genes among the isolates analyzed in this study (P > 0.05).
Staphylococcus is the most prevalent pathogen in nosocomial bacteremia and sepsis (1). The ability of this organism to cause persistent infection has been attributed mainly to biofilm formation (4). In the present study we analyzed biofilm formation by 127 staphylococcal isolates from true and false bacteremia specimens from patients in a Brazilian tertiary hospital to clarify whether a combination of methods could help in determining the clinical relevance of the isolates.
Biofilm production by staphylococci has been evaluated mainly by MPA, the gold standard method, and by CRA, a simple and inexpensive test, but the results have been variable. Furthermore, only a few studies have used more than one phenotypic method to analyze this virulence factor. Grinholc and coworkers showed that among 48 ica genes positive S. aureus isolates from bacteremia, 50% and 46% produced biofilm on CRA and MPA, respectively (25). A previous study analyzed 15 similar isolates from bacteremia but only one isolate was found to be MPA positive (8). On the other hand, Satorres & Alcaráz verified that all nine ica genes positive S. aureus isolates from blood cultures were also positive on CRA (11), a finding similar to that of Gad et al., who analyzed 18 S. aureus isolates from patients’ urinary tract catheters and detected a correlation of 100% between the presence of ica genes and biofilm production on CRA and MPA (9). In the present study we found a similar rate of positivity for both CRA (67.5%) and MPA (62.5%) in 40 S. aureus isolates from blood cultures that were positive for the icaA gene. Interestingly, this figure increased to 85% (P= 0.022 in relation to MPA and P= 0.066 in relation to CRA) when the results of both phenotypic methods were combined, making the correlation with the presence of the ica gene closer. Moreover, since negative isolates for the ica gene were also negative for both phenotypic methods analyzed, we think that a combination of methods would more accurately predict the presence of this gene in S. aureus isolates from blood cultures.
Regarding S. epidermidis, through the CRA method, Růzicka et al. (6) and Satorres & Alcaráz (11) found that 65% and 100% of positive isolates from blood were ica genes positive, respectively. Another study that analyzed 65 ica gene positive S. epidermidis isolates from infected orthopedic devices showed that 100% were positive for CRA (15). Among the studies that have employed the two phenotypic methods used in the present work, one detected 100% positivity by both methods in 44 ica genes positive S. epidermidis isolates from blood (5). Koskela et al. found that the rate of positivity was higher in MPA (81%) than in CRA (44%) among 16 S. epidermidis isolates carrying ica genes obtained from prosthetic joint infections (26). Our results were similar to described by these authors: we found 66.7% for MPA and 41.7% for CRA among 12 S. epidermidis isolates positive for ica genes, increasing to 75% when we used a combination of both methods. It is important to emphasize that, among the S. epidermidis isolates positive for ica genes and MPA and/or CRA, the frequency was higher for isolates from true bacteremia (100%) than for those from false bacteremia (40%). Additionally, among biofilm producing S. epidermidis isolates, seven (77.8%) were from true bacteremia and four of them were simultaneously ica genes, MPA and CRA positive, showing the importance of the combination of methods in confirming biofilm formation and determining the clinical relevance of S. epidermidis in blood cultures.
Some authors have shown that addition of glucose increases the detection of biofilm-producer isolates in S. aureus (27) and S. epidermidis (28) species. In our study, we verified an increased identification of biofilm-production by the addition of this component to the CRA (from 15% to 23.4%). However, a period of incubation at room temperature to a total of 72 hr on CRA with glucose contributed more significantly to the rate of detection of biofilm producer S. aureus isolates (67.5% detected, P < 0.05).
The detection of resistance of staphylococci to oxacillin is important in guiding therapy and preventing unnecessary treatment with vancomycin (1). In this study, the mecA gene was detected in all but one resistant isolates of S. aureus by the disk diffusion test. Among all CNS isolates two were false resistant and one was false sensitive by this method, although all isolates had been correctly detected by the agar screening test with 4 μg/mL oxacillin, a concentration previously found by us to be effective in detection of methicillin resistant CNS isolates (23).
In conclusion, according to this study, combination of the CRA and MPA methods provides more accurate prediction of the presence of the ica gene in S. aureus isolates and helps to confirm biofilm production in S. epidermidis isolates, contributing to determining their clinical relevance in blood cultures.
We thank Dr. Andréa Gonçalves Antonio and Dennis de Carvalho Ferreira for critical reading of the manuscript. This study was supported by grants from: Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ, Carlos Chagas Filho Foundation for the Support of Research in the State of Rio de Janeiro), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, National Council of Scientific and Technological Development), Coordenação de Aperfeiçoamento Pessoal de Nível Superior (CAPES, Coordination of Personal Perfection at a Superior Level), Fundação Universitária José Bonifácio (FUJB, José Bonifácio University Foundation), and Programa de Núcleos de Excelência (PRONEX, Program of Nuclei of Excellence).
- 3National Nosocomial Infections Surveillance System. (2004) National Nosocomial Infections Surveillance (NNIS) system report, data summary from January 1992 to June 2004, issued October 2004. Am J Infect Control 30: 458–75.
- 21Clinical and Laboratory Standards Institute. (2009) Performance Standards for Antimicrobial Disc Susceptibility Test, 10th edn. Approved Standards: M02-A10. Wayne , PA , USA : CLSI.
- 22Clinical and Laboratory Standards Institute. (2009) Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically, E8th edn. Approved Standard M07-A8. Wayne , PA , USA : CLSI.
- 252007) Evaluation of biofilm production and prevalence of the icaD gene in methicillin-resistant and methicillin-susceptible Staphylococcus aureus strains isolated from patients with nosocomial infections and carriers. FEMS Immunol Med Microbiol 50: 375–9., , (