INTRODUCTION
Infections during the neonatal period lead to high levels of morbimortality.Sepsis is one of the most relevant complications of neonatal infections, and it increases the hospital care required for newborn babies1–3. Coagulase-negative staphylococci (CoNS), which are part of the normal skin microbiota, are the most frequently isolated pathogens in blood stream infections. In particular, Staphylococcus epidermidis is capable of causing infections in both immunocompromised patients and newborns4–8.
CoNS infections are frequently associated with the use of invasive medical devices, and biofilm formation is a major virulence determinant in this group of bacteria4–7–11. The formation of biofilms allows these microorganisms to adhere to biomaterials, and biofilm formation is related to infection recurrence and therapy failure, as the organisms within biofilms are more resistant to antimicrobials and host defenses4–7–9,11,12.
In recent decades, there is increasing awareness of the great difficulty in eradicating infections caused by biofilm-forming bacteria, since these microorganisms are more resistant to and can survive in the presence of various antimicrobials, requiring removal of the related medical device or surgical methods to cure the infection6,12–15.
CoNS develop into biofilms via a complex, multifactorial process that can be divided into four phases: adhesion, accumulation, maturation, and detachment, each involving specific molecular factors5,13,16. Studies suggest that polysaccharide intercellular adhesin (PIA) is the most important component of Staphylococcus spp. biofilms. Biofilm production in Staphylococcus spp. is mediated by the chromosomal intracellular adhesion (ica) gene products, which are involved in cell adhesion and are arranged in an operon structure containing four biosynthesis genes icaABCD7,11,13.
Considering the large number of infections caused by biofilm-forming bacteria, early detection and eradication of these microorganisms is necessary. A phenotypic method, along with detection of the genes in the ica operon, is generally used to verify biofilm production in Staphylococcus spp. infections7,12. The purpose of the present study was to assess four methods for the detection of biofilm production in CoNS isolates from blood cultures of newborn patients in a tertiary hospital in southern Brazil. The four methods evaluated were tissue culture plate (TCP), the tube method (TM), congo red agar (CRA), and polymerase chain reaction (PCR). This is the first study to assess biofilm production in CoNS isolates from blood cultures of newborn patients at this hospital.
METHODS
Location and duration of study
This study was conducted at the Laboratório de Bacteriologia, Departamento de Análises Clínicas e Toxicológicas, Centro de Ciências da Saúde of Universidade Federal de Santa Maria, Santa Maria, Rio Grande do Sul, Brazil. Samples were isolated at a tertiary hospital in southern Brazil during a one-year period (2014).
Isolates
One hundred and seventy-six (176) CoNS isolates were obtained from the blood cultures of newborns. Blood cultures were required whenever a significant clinical sign was present, such as increased body temperature (>38°C), hypothermia (<36°C), leukocytosis (>10.000 leukocytes/mm3, especially with left deviation), or absolute granulocytopenia (<1000 leukocytes/mm3). For the cultures, a 3mL sample of blood was collected from each newborn and placed in a blood culture bottle. At least two blood cultures were used for each sample. Isolates were processed with the BACTEC 9240® automated blood culture system (Becton Dickson, Sparks, MD), and samples were considered positive when two or more blood cultures developed CoNS. Species-level identification of CoNS and antimicrobial sensitivity profiles were performed with the Vitek® 2 system (bioMérieux, France). The sensitivity profiles were determined according to the Clinical and Laboratory Standard Institute (CLSI) guidelines17 at the time of the study.
Blood cultures in which CoNS growth was detected were subsequently sent to the Laboratório de Bacteriologia, Departamento de Análises Clínicas e Toxicologicas, Centro de Ciências da Saúde at UFSM, where they were inoculated on tryptone soya agar (TSA) and incubated at 35 ± 2°C for 18-24 h. The colonies were stored in tryptone soya broth (TSB) containing 15% glycerol at −80°C until biofilm detection.
PCR, TM, and CRA were performed in duplicate, and TCP was performed in quadruplicate. S. epidermidisAmerican Type Culture Collection (ATCC) 12228 and Staphylococcus aureus ATCC 25923 (a strong biofilm producer) were used as the negative and positive controls, respectively.
Biofilm detection by the genotypic method (PCR)
DNA was extracted by the boiling method of Pérez-Pérez and Hanson18. The 16S rRNA gene sequence was determined to verify the presence of CoNS bacterial DNA, and then the icaACD genes were detected by PCR, which is considered the gold standard for detecting biofilm formation ability14,19. The primers used, which are shown in Table 1, were obtained from GBT Oligos® and were designed based on the Staphylococcus epidermidis icaACD sequence (GenBank accession number U43366.1). To amplify the icaACD genes, a Thermo Cycler (model 2720, Biosystems) was used and programmed with the following cycling conditions: an initial step at 94°C for 5 min followed by 50 cycles of 30 sec each at 94°C, 55.5°C, and 72°C, with a final step at 72°C for 1 min. Each PCR simplex reaction contained 17.05 µL of ultrapure water, 1.75 µL of buffer (10×; Ludwig-Biotec®), 0.75 µL of MgCl2 (50 mM; Ludwig-Biotec®), 2 µL of dNTPs (100 mM; Ludwig-Biotec®)s 0.2 µL of Taq DNA polymerase (5U/mL; Ludwig-Biotec®), 0.625 µL of forward primer, 0.625 µL of reverse primer, and 2 µL of DNA. The DNA fragments were analyzed by 1.5% agarose gel electrophoresis, and the bands were visualized with a photo documentation system (KODAK DC 290, using 1D software, version 3.6).
Gene | Primer | Sequence (5ʹ-3ʹ) | Product size (bp) |
---|---|---|---|
16S rRNA | 16S-F | CCTATAAGACTGGGATAACTTCGGG | 791 |
16S-R | CTTTGAGTTTCAACCTTGCGGTCG | ||
icaA | icaA-F | ACAGTCGCTACGAAAAGAAA | 103 |
icaA-R | GGAAATGCCATAATGACAAC | ||
icaC | icaC-F | TAACTTTAGGCGCATATGTTTT | 400 |
icaC-R | TTCCAGTTAGGCTGGTATTG | ||
icaD | icaD-F | ATGGTCAAGCCCAGACAGAG | 198 |
icaD-R | CGTGTTTTCAACATTTAATGCAA |
Congo red agar (CRA)
Isolates were analyzed according to the method described by Freeman et al.20. Briefly, isolates were inoculated on sheep blood agar, incubated at 35 ± 2°C for 24 h, and then transferred to CRA and incubated at 35 ± 2°C for 24 h. Biofilm-producing isolates formed black colonies, whereas non-biofilm-producing isolates formed red colonies.
Tube method (TM)
Isolates were assessed by the TM as described by Christensen et al.21. They were inoculated on sheep blood agar and incubated at 35 ± 2°C for 24 h. Then, selected colonies were inoculated into test tubes containing 2.0 mL of TSB and incubated at 35 ± 2°C for 48 h. Later, the contents were removed with a pipette, and 1.0 mL of a 0.4% aqueous solution of trypan blue was added to each tube. The stain was removed with a pipette and after a 1 min incubation, the results were read by visual observation. If colored bacteria adhered to the tube wall, the isolate was considered to be a biofilm producer, whereas if no colored microorganisms adhered to the tube wall, the isolate was considered to be a non-biofilm producer.
Tissue culture plate (TCP)
Isolates were assessed by the TCP method as described by Christensen et al. (1985)22,with some modifications. The bacteria were inoculated on sheep blood agar and incubated at 35 ± 2°C for 24 h. Then, colonies were transferred to test tubes containing 2.0 ml of TSB and incubated at 35 ± 2°C for 24 h.
Using multichannel pipettes, 200 µL of TSB containing 1% glucose was added to each well of a 96-wellflat bottom polystyrene tissue plate. Then, a 200 µL aliquot of the aforementioned bacterial suspension in TSB was added to each well; each isolate was inoculated into four parallel wells, and the plates were incubated at 35 ± 2°C for 24 h. Then, the contents of each well were aspirated, and the wells were washed four times with 200 µL of phosphate-buffered saline (PBS, pH 7.2). The wells were stained with 100 µL of 2% crystal violet for 1 min.Then, the stain was aspirated, and wells were washed with distilled water. Plates were dried for 1 h at room temperature. The optical density (OD) at 570 nmwas then read with an EpochTM Multi-volume Spectrophotometer (Biotech)22.
Sterile TSB containing 1% glucose, which underwent all procedures described above, without the addition of microorganisms, was included as a medium control. A cut-off value for the detection of biofilm formation was calculated as the standard deviation multiplied by three plus the OD of the medium control (cut-off value = standard deviation × 3 + average OD medium control)22.
Isolates were classified into three categories: non-adherent (OD less than or equal to the cut-off value,i.e., OD≤ 0.116), weakly adherent (OD >0.116 and ≤0.232), and strongly adherent (OD> 0.232).
Statistical analysis
Sensitivity and specificity were calculated based on comparison of the phenotypic methods and PCR, which is considered a gold standard. The Kappa index (k) was also calculated to verify agreement between the results obtained from the different methods7,23. Data were analyzed using the Statistical Package for Social Sciences (SPSS) 20.0 for Windows. A p value less than 0.05 was considered statistically significant, and a 95% confidence interval (CI) was used.
RESULTS
The 16S rRNA gene was amplified in all 176 CoNS isolates, and 41.5% (73/176) were positive for icaA, icaC, or icaD by PCR; 30.1% (53/176) carried icaACD, 11.4% (20/176) carried icaAD, and none of the isolates expressed icaA, icaC, or icaD alone (Figure 1).
On CRA, 42% (74/176) of the isolates showed black-colored colonies, indicative of biofilm production, and 58% (102/176) showed red-colored colonies, indicating no biofilm production (Figure 2). The TM showed that 38.6% (68/176) of isolates adhered to the tube walls and were considered positive for biofilm production, while 61.4% (108/176) of isolates did not adhere to the tube walls and were considered negative for biofilm production (Figure 3). In the TCP assay, 40.9% (72/176) of isolates produced biofilms; 21% (37/176) were weakly adherent, and 19.9% (35/176) were strongly adherent. Of the weakly adherent isolates, 78.4% (29/37) expressed icaAD, and 21.6% (8/37) expressed icaACD. All strongly adherent isolates (35/35) expressed all three genes (icaACD).
In our study, CRA showed 79% sensitivity (95% CI, 70-88), 84% specificity (95% CI, 76-92), 82% accuracy, and good agreement (Kappa = 0.64); the TM assay showed 78% sensitivity (95% CI, 69-87), 89% specificity (95% CI, 82-96), and 85% accuracy, and good agreement (Kappa = 0.83); and the TCP method showed 99% sensitivity (95% CI, 97-100), 100% specificity (95% CI, 99-100), 99% accuracy, and excellent agreement (Kappa = 0.99), when compared to the gold standard, PCR.
Among the isolates that were positive for icaACD or icaAD by PCR, S. epidermidis was the prevalent species (67.1%, 49/73), followed by S. warneri (8.2%, 6/73). Table 2 shows the number of biofilm-producing CoNS species that were positive by the four tested techniques (PCR, CRA, TM, and TCP). Table 3 shows the percentages of biofilm-producing and non-biofilm-producing isolates that were resistant to each tested antimicrobial. All isolates in this study were sensitive to linezolid, tigecycline, and vancomycin.
Species | PCR | CRA | TM | TCP |
---|---|---|---|---|
S. epidermidis | 49 | 52 | 48 | 48 |
S. warneri | 6 | 5 | 6 | 6 |
S. capitis | 5 | 7 | 4 | 5 |
S. haemolyticus | 4 | 3 | 3 | 4 |
S. hominis | 3 | 2 | 2 | 3 |
CoNS | 2 | 2 | 1 | 2 |
S. lugdunensis | 1 | 1 | 1 | 1 |
S. saprophyticus | 1 | 1 | 1 | 1 |
S. xylosus | 1 | 1 | 1 | 1 |
S. cohnii | 1 | – | 1 | 1 |
Total | 73 | 74 | 68 | 72 |
PCR: polymerase chain reaction; CRA: congo red agar; TM: tube method; TCP: tissue culture plate; CoNS:coagulase-negative staphylococci.
Antimicrobials | n | Resistant | n | Resistant |
---|---|---|---|---|
biofilm producers | non-biofilm producers | |||
n = 73 | n = 103 | |||
Ciprofloxacin | 51 | 69.9% | 41 | 39.8% |
Clindamycin | 30 | 41.1% | 10 | 9.7% |
Erythromycin | 44 | 60.3% | 42 | 40.8% |
Gentamicin | 45 | 61.6% | 41 | 39.8% |
Oxacillin | 71 | 97.3% | 62 | 60.2% |
Penicillin | 73 | 100% | 49 | 47.6% |
Trimethoprim/ Sulfamethoxazole | 41 | 54% | 31 | 30.1% |
*Biofilm production was assessed with the gold standard technique (PCR).
DISCUSSION
In this study, we analyzed the biofilm-formation ability of 176 CoNS isolates by four different techniques. Among the 176 isolates, 41.5% of were positive for icaA, icaC, or icaD by PCR; 30.1% expressed icaACD, and 11.4% expressed icaAD. Oliveira and Cunha7 compared different methods for detecting biofilm formation in CoNS isolates from clinical specimens of newborns and the nasal cavity of healthy people in Botucatu, São Paulo, Brazil. They found that 82% of isolates were positive for icaA, icaC, or icaD by PCR; 40% carried icaAD, and 42% carried all three genes (icaACD). These percentages were higher than those in our study. Zalipour et al.12also assessed this virulence determinant in Staphylococcus spp. obtained from different clinical specimens collected in two school hospitals in Iran, and also found a higher rate than that in our study, as 81.9% of the S. epidermidisisolates expressed icaAD.
Regarding the phenotypic methods, 42% of the isolates showed biofilm production by the CRA assay. Similar results were reported by Öcal et al.11for CoNS isolates in samples from nasal cavities, catheters, and blood cultures of patients admitted to a hospital in Turkey; 40.3% of isolates produced biofilms by the CRA technique. In contrast, Oliveira and Cunha7 and Zalipour et al.12 reported higher rates of biofilm formation on CRA, with73% of CoNS and 70.8% of S. epidermidis, respectively. Hassan et al.24 assessed microorganisms isolated from different clinical materials in a hospital in Pakistan, and only 10% of the tested bacteria were biofilm producing using the CRA technique.
In this study, 38.6% of isolates were considered positive for biofilm formation by the TM. Hassan et al.24 and Oliveira and Cunha7found somewhat higher values using TM, with ~49% and 82% biofilm-producing isolates, respectively.
Using TCP, 40.9% of our isolates were determined to be biofilm producing; 21% were weakly adherent, and 19.9% were strongly adherent. Oliveira and Cunha7observed higher indices, and 81% of the isolates were biofilm producing by TCP; 35% were weakly adherent, and 46% were strongly adherent. Rani et al.25 assessed biofilm formation in S. epidermidis obtained from clinical specimens in a hospital in India using TCP, and all isolates produced biofilm. In contrast, Hassan et al.24reported a value closer to that in our research, with 54.5% biofilm-producing isolates.
In the present study, of the weakly adherent isolates, 78.4% carried icaAD and 21.6% carried icaACD, whereas all the strongly adherent isolates carried all three genes. None carried one ica gene alone. Oliveira and Cunha7reported that only 56.5% of the strongly adherent isolates contained all three genes (icaACD). These studies demonstrate that the presence of the icaACD genes is an important virulence determinant in clinical isolates of CoNS, since their expression is associated with the production of PIA, a major component of biofilms in Staphylococcus spp.7. We assumed that the presence of only one of the genes in the ica operon is not sufficient for biofilm formation26.
The expression of icaA alone induces low enzymatic activity, as these strains typically produce only small amounts of PIA. The presence of icaD favors the production of polysaccharide, and larger chains are synthesized in the presence of icaC. The expression of icaAD or icaACD stimulates increased biofilm production; whereas icaB appears to function as a deacetylase for the formation of PIA7,11,13,27. In addition to this ica operon-related mechanism of biofilm formation, biofilms can also be formed through protein adhesion mediated by surface proteins, such as biofilm-associated protein (Bap), its homologs Bhp and accumulation associated protein (Aap), and extracellular matrix binding protein (Embp), via cell adhesion and accumulation28–30.
In our study, the CRA assay showed 79% sensitivity, 84% specificity, and 82% accuracy (kappa = 0.64); the TM assay showed 78% sensitivity, 89% specificity, and 85% accuracy (kappa = 0.68); and the TCP assay showed 99% sensitivity, 100% specificity, and 99% accuracy (kappa = 0.99), when compared to the gold standard, PCR. Oliveira and Cunha7 compared the three techniques used in our study (CRA, TM, and TCP) to the gold standard technique (PCR) for strains that carried the biofilm-producing genes icaACD. They observed 100% sensitivity and specificity for the TM, 89% sensitivity and 100% specificity for the CRA assay,and 96% sensitivity and 94% specificity for the TCP assay7. Thus, the authors concluded that the TM is the best method for biofilm detection due to its high sensitivity and specificity7. In our study, the CRA and TM assay failed to show consistent results. Thus, we concluded that TCP is the most suitable method for routine laboratory testing.
According to some researchers, the CRA assay is easier to implement than other methods; however, it is not recommended for biofilm detection in routine laboratory testing since it can yield false positive or false negative results7,31. In fact, a study conducted in India showed that the TM cannot be used as a general screening test to identify biofilm-producing strains32. However, Oliveira and Cunha7 recommended this qualitative method (TM) for biofilm research because it is a low cost method that yields reliable results. Although the TCP method is now considered the gold standard by some researchers24,25,32, as it is time-consuming, it is not widely used in routine laboratories7.
Among the biofilm-producing CoNS species detected by the standard technique (PCR), S. epidermidis was the most prevalent, at 67.1%, followed by S. warneri (8.2%). Oliveira and Cunha7reported S. epidermidis as the most prevalent species, which was present at a higher rate than in our study7. The fact that S. epidermidis shows the highest rate of biofilm production capacity may be due to special mechanisms, such as their ability to colonize the surface of medical devices, resilience, and because they colonize the skin of babies as soon as they are born. Furthermore, among CoNS, S. epidermidis is the species most frequently isolated from neonatal infections, and it is frequently detected in infections associated with invasive devices7,13,15,25.
In this study, high rates of resistance to penicillin, oxacillin, and clindamycin were detected among the biofilm-producing isolates. Research has shown that biofilm formation is associated with an increase in microbial resistance, and the acquisition of resistance characteristics via gene transfer is possible within biofilms33–35.
Our results showed that approximately 42% of CoNS isolates were positive for icaA, icaC, or icaD by PCR, and strains containing all three genes (icaACD) showed a greater capacity to form biofilms. Regarding the phenotypic methodologies, the TM and CRA assay presented low of sensitivity and specificity values, and are not recommended for routine laboratory use. Thus, although the quantitative TCP method is a time-consuming technique, it would be ideal for verifying biofilm production in nosocomial CoNS isolates.
This study has some limitations. First, even in the presence of ica genes, a strain may not form a biofilm in vitrodue to non-expression of these genes. Conversely, some isolates may produce biofilm, even in the absence of the ica genes, which could interfere with the sensitivity and specificity. In addition, it is known that biofilm production can be induced via other ica-independent genes, such as Bap, Bhp, Aap, and Embp, which were not examined in this study.