Antibiotic Resistance and Biofilm Production in Staphylococcus epidermidis Strains, Isolated from a Tertiary Care Hospital in Mexico City

Staphylococcus epidermidis strains isolated from nosocomial infections represent a serious problem worldwide. In various Mexican states several reports have shown isolates from hospitals with antibiotic resistance to methicillin. In Mexico City, there is scarce information on staphylococcal infections in hospitals. Here, our research findings are shown in a four-year period study (2006–2010) for Staphylococcus epidermidis strains. Susceptibility and/or resistance to antibiotics in SE strains were assessed by phenotypic and molecular methods as mecA gene by PCR, as well as the correlation with biofilm production for these isolates and the relationship to the infection site. Out of a total of 161 (66%) negative biofilm SE strains, just 103 (64%) SE strains were confirmed as MRSE by PCR to mecA gene. From 84 (34%) positive biofilm SE strains, 76 (91%) were confirmed as MRSE by PCR to mecA gene. Higher percentages of resistance to antibiotics and higher number of resistance markers were found in biofilm-forming clinical strains (9 to 14) than non-biofilm-forming SE strains (3 to 8). These research findings represent a guide to establish infection control programs for this hospital.


Introduction
Staphylococcus epidermidis (SE) is a saprophyte which is part of the normal human skin and mucous membranes micro�ora. Furthermore, SE is one of the most common etiologic agents of nosocomial infections (NIs) worldwide. Moreover, SE is the most prevalent etiology of NI in pediatric intensive care units (ICUs) in Mexican hospitals [1,2].
However, SE is primarily associated with infections in patients implanted with medical devices, such as prosthetic heart valves and orthopedic prostheses, mainly in immunocompromised children. Furthermore, SE was isolated with higher incidence to catheter-related bacteremia in a report from secondary care Hospital in Durango, México [3,4].
�io�lm is mainly made of polysaccharide component; it seems to be the most important factor by which SE adheres to and coloni�es arti�cial materials (catheters) commonly implanted in patients with NI. �io�lm is believed to make clinical SE strains more resistant to administered antibiotics and to host defense mechanisms and highly contributed to cause NI in patients [2,[5][6][7][8][9].
In clinical practice, SE has become one of the most significant species among methicillin-resistant coagulase negative staphylococci (CoNS). ere are various scienti�c world reports that stated that approximately between 80% and 90% of SE strains isolated from patients with NI carried the mecA gene [10][11][12][13]. e presence of mecA gene seems to be enhanced in bio�lm producers and enable SE strains to show increasing resistance to different groups of antibiotics [14,15].
In a university hospital at Monterrey, Mexico, staphylococci methicillin-resistant strains were detected by molecular methods, and SE was the most common species identi�ed [16]. A general hospital of Mexico City performs a series of molecular tests to determine icaA gene and consequently bio�lm production from staphylococci isolates and research �ndings here show that SE bio�lm producers were the most common species found [17].
Actually, there are only few hospital reports that express the genuine situation of SE infections in Mexico, which means that more research is required to understand the real scenario about NI in other hospitals, especially in Mexico City, which is one of the biggest and crowdest cities in the world. e aim of the present study is the phenotypic and genotypic characterization related to bio�lm production, methicillin and other antibiotic resistance patterns to SE strains isolated from NI in a tertiary-care-level hospital in Mexico City.
2.1. mecA PCR Assay. PCR assay was performed for all 245 SE strains searching an ampli�cation product of 458 bp that represent a fragment of mecA gene. e PCR primers used were 5 ′ -ATGGCAAAGATATTCAACTA-3 ′ (upstream) and 5 ′ -GAGTGCTACTACTCTAGCAAAGA-3 ′ (downstream). e primers were designed for mecA gene by our group using a Primer3 free soware (http://frodo.wi .mit.edu/primer3/) and the information available in GenBank/NCBI (http://www.ncbi.nlm.nih.gov/) to genome sequence for SE strain RP62A (accession NC_002976) the primer sequences for mecA gene were compared with other staphylococci genomes using the BLAST algorithm (http://www.ncbi.nlm.nih.gov/BLAST) then our PCR primers were considered speci�c to this gene.
Ten to twenty colonies of each SE isolate grown on BHI agar plates were suspended in 250 L of lysis buffer solution (0.01 M Tris-HCL SIGMA-ALDRICH, Co. St. Louis MO; 0.01 M EDTA SIGMA-ALDRICH, Co. St. Louis MO; lysozyme (200 g/mL; SIGMA-ALDRICH, Co., St. Louis MO) and lysostaphine (20 g/mL; Sigma-Aldrich, Co., St. Louis MO)); this bacterial suspension was incubated at 37 ∘ C during 90 minutes then aer the samples were warmed . e experimental conditions used were as follows: an initial denaturation phase at 94 ∘ C for 10 minutes, then 30 cycles of 94 ∘ C for 30 seconds, annealing at 49.7 ∘ C for 30 seconds, and extension at 72 ∘ C for 30 seconds, with a �nal primer extension at 72 ∘ C for 10 min. e PCR products were separated by electrophoresis through 1% agarose gels in 1X TAE buffer (40 mM Tris Acetate, 1 mM EDTA; Invitrogen Life Technologies Carlsbad, CA, USA) run the gels for 40 minutes at 95 Volts. e gels were stained with 0.08 L/mL of ethidium bromide (10 mg/mL; Invitrogen Life Technologies Carlsbad, CA, USA), and bands were visualized under UV light.
2.2. �i���m �es�. e bio�lm production for SE clinical isolates was assessed by Congo Red Agar plates assay (CRA). is culture medium (CRA) was prepared with Brain Heart Infusion agar (BHI) (OXOID LTD., Basingstoke, Hampshire, England) supplemented with Congo Red dye; 0.8 g/L (Sigma-Aldrich, Co., St. Louis, MO) and sucrose; 36 g/L (J. T. Baker, México). All clinical SE strains to be tested for bio�lm formation were inoculated on CRA plates. ese CRA plates were incubated at 35 ∘ C for 24 hrs then aer the plates were maintained at room temperature for 24 hrs. CRA plates were checked for black colonies which represented the SE bio�lm producers or positive (B+), in contrast to the red colonies, which represented the SE non-bio�lm producers or negative (B−) [18].

Results
From 245 SE clinical isolates, 34% (84) were detected as black colonies or bio�lm-forming strains on CRA plates. e remaining SE strains 66% (161) were grown as red colonies on these plates and therefore were considered as non-bio�lm producers or negative strains (Figures 1(a) and 1(b)). Patterns of susceptibility/resistance to antibiotics of a total of 245 SE strains were performed by automated Micro Scan system. SE clinical strains of bio�lm producers were displayed lightly more resistant to 14 antimicrobial agents tested than SE strains of non-bio�lm producers (Table 1).
Higher percentages of resistance to antibiotics and higher number of resistance markers were found in bio�lm producer mecA (+) clinical strains (9 to 14) than non-bio�lm producer mecA (−) strains (3 to 8) (Figures 2 and 3).
e percentage of antibiotic resistance in SE bio�lm producer strains is more relevant in the 40% of the mecA gene positive strains mainly in those that present resistance to 13 resistant markers. In mecA negative strains are appreciated for strains with 6 resistance markers corresponding to 25% of strains ( Figure 3).

Discussion
In the present study we observed a quite good correlation in antibiotic resistance and bio�lm production for clinical MRSE strains from patients with NI at this tertiary care hospital in Mexico City. e majority of these clinical SE strains were positive for molecular detection of the mecA gene by PCR. Data from other hospitals in Mexico, such as one report by Diemond et al. [17], determined the presence of bio�lm in SE and CoNS strains for icaA operon by PCR method. In this study SE bio�lm producer SE strains were found in higher percentages in catheter sites. In our study 90 out of 245 strains were isolates from catheter site, and 37 out of 90 strains were positive for bio�lm production. Calderón et al. [19] in a pediatric hospital of Mexico City found resistance in 53.4% of CoNS strains to oxacillin. Other groups such as Garza et al. [16], in a tertiary care hospital, made a surveillance study for the distribution of SCCmec cassettes and its relationship to MRCoNS strains; they found that SE was the most common species identi�ed and carrying the cassettes type III, IVa, and V for methicillin resistance. In a tertiary care hospital a study was performed by Kato-Maeda et al. [20] group, describing the resistance to antibiotics in some isolates causing bacteremia during a six-year study period (1995)(1996)(1997)(1998)(1999)(2000), SE was the third most frequently isolated bacteria and the oxacillin resistance pattern was increased from 41 % to 61% during this period [8,9].
Our research �ndings show that MRSE strains carry a higher number of resistance determinants in bio�lmproducer strains. ese strains were more frequently isolated in clinical samples from intravascular catheters in this tertiary care hospital.
Pro�les of resistance to 17 antibiotics tested were slightly higher in SE strains of bio�lm producers than in nonproducer strains. e phenotypic bio�lm characterization by CRA plates assay demonstrated a simple, rapid, cheap, and reliable approach for detection of bio�lm producers in SE isolates from patients with staphylococci NI. It is highly recommended for the future a validation of this test versus genotypic test such as PCR detection for icaA and icaD genes. is molecular PCR methodology will be soon applied in this hospital for detection of other relevant antibiotic resistance markers, in order to establish a permanent NI-control program.
e PCR mecA gene detection assay meets our requirements as a reliable molecular method for characterization of methicillin-resistant SE strains. Both methods are urgently needed for molecular epidemiology studies and outbreak control in Mexican hospitals. Finally our data suggest that this methodology could be used to establish infection control procedures for this tertiary care hospital in order to prevent nosocomial infections.