Bacterial isolates

Bacterial isolates were obtained from sputum samples collected from 50 pediatric patients attending the CF clinic of the Hospital Infantil de México Federico Gómez (HIMFG), Mexico City, Mexico, from August 2016 to January 2018⁴. Parents of children with CF who agreed to participate in this study signed a consent form authorizing the collection of sputum samples to identify bacterial pathogens. Patient samples were collected as part of routine hospital care. Each patient had appointments scheduled every 3 to 6 months, although some patients and their parents do not always attend their scheduled appointments at HIMFG.

Sputum samples or cough swabs were transported to our laboratory for processing. Sputum samples were dissolved weight/volume (1:1) in sputolysin (Merk-Millipore, Darmstadt, Germany) for 30 minutes at 37^oC. Dissolved sputum samples and cough swabs were used to inoculate salt and mannitol, chocolate, blood, MacConkey (DIBICO, State of Mexico, Mexico), and cetrimide (Becton Dickinson, New Jersey) agar media plates. Plates were incubated at 37^oC for 24 hours. Chocolate agar and blood agar plates were also incubated at 37^oC for 24 hours under microaerophilic conditions. Bacterial isolates were identified by standard microbiological methods⁴. All samples were stored at −70^oC until analysis.

Pulsed-field gel electrophoresis

The bacterial relatedness of 50 clinical isolates of S. aureus was determined by pulsed-field gel electrophoresis (PFGE). Bacterial genomic DNAs were purified and prepared as described elsewere¹⁴. Genomic DNAs were digested with SmaI (Invitrogen) for 24 h and resolved by PFGE using a Gene Path system (BioRad^®USA). Bacterial relatedness among S. aureus clinical isolates was determined according to the Tenover criteria and the use of the Dice coefficient, as previously described⁴. An isolate was considered a member of the same pulsotype when it had a > 85% correlation.

Antibiotic susceptibility testing

S. aureus isolates were tested for antimicrobial susceptibility to ciprofloxacin, levofloxacin, moxifloxacin, gentamicin, tigecycline, trimethoprim/sulfamethoxazole, oxacillin, erythromycin, clindamycin, linezolid, vancomycin, tetracycline, and rifampicin. The minimum inhibitory concentrations (MICs) were determined using the VITEK^®2 system (bioMérieux^®SA). Methicillin resistance was determined using the cefoxitin test by the disk diffusion method (Kirby-Bauer). Quality control and interpretation of results were performed according to Clinical and Laboratory Standards Institute (2019) guidelines¹⁵. S. aureus ATCC 43300, 25923, and USA300 were used as standard quality controls. The multidrug resistance (MDR) phenotype is defined as non-susceptibility to ≥ 1 agent from ≥ 3 antimicrobial categories¹⁶.

DNA extraction and identification of mecA and lukPV genes by polymerase chain reaction (PCR)

S. aureus isolates were grown on salt and mannitol agar plates for 24 hours at 37^oC. An isolated colony was resuspended in 100 mL of MilliQ water and boiled for 10 minutes. The bacterial suspension was centrifuged at 10,000 rpm at 4^oC for 10 minutes. Supernatants containing genomic DNA were used for PCR reactions. Primers mecA(F): 5’-TGGCTATCGTGTCACAATCG-3’ and MecA(R): 5’-CTGGAACTTGTTGAGCAGAG-3’ were used for mecA amplification, whose amplification product is 310pb¹⁷. The lukPV gene was amplified with primer LukPV(F): 5’-ATCATTAGGTAAAATGTCTGGACATGATCCA-3’ and LukPV(R): 5’-GCATCAACTGTATTGGATAGCAAAAGC-3’, with an amplification product of 433pb¹⁸.

PCR was performed as follows: we mixed 10 mM of each primer, 1X GoTaq^® Green Master Mix, and 3 mL of DNA template extracted by boiling. The conditions for gene amplification were one cycle at 94^oC for 5 min, 30 cycles at 94^oC for 30 s, 55^oC for 30 s, and 72^oC for 90 s with a final extension step at 72^oC for 5 min. PCR products were resolved on a 1% (w/v) agarose gel for one hour at 100 volts. The gel was stained with ethidium bromide (1 mg/mL) for 5 min and washed twice in deionized water. The gel was analyzed on a UV light transilluminator using Quantity One software (BioRad^® USA).

Reagents, cells, and growth conditions

Fetal bovine serum (FBS) and RPMI-1640 cell culture media were obtained from Invitrogen, PBS and Luria-Bertani (LB) broth from Sigma-Aldrich, and salt and mannitol agar from DIBICO (State of Mexico, Mexico). Human THP1 cells were obtained from ATCC^® TIB-202™; 3x10⁵ human THP1 cells were differentiated into macrophages (THP1m) using PMA 100 ng/mL (Sigma-Aldrich) for 24 hours¹⁹. Prior to infection, the medium was changed to RPMI medium with no antibiotics. THP1m cells monolayers were infected with S. aureus at an MOI of 50. To synchronize the infection, we centrifuged the cells at 1,200 rpm for 1 min and then incubated the plates for one hour. After infection, the cells were washed three times with PBS to remove extracellular bacteria; infected cells were returned to the incubator for an additional 24 h in RPMI medium supplemented with gentamicin (100 mg/mL).

Cytotoxicity assays

Supernatants from THP1m and THP1m colonized with S. aureus were used to quantify cytosolic enzyme activity of lactate dehydrogenase (LDH; Promega, Madison WI, USA). The following formula determined the percentage of LDH activity:

Presence and genotyping of S. aureus isolates from CF pediatric patients

Of the 50 pediatric patients studied, the most frequently isolated bacterial pathogen was P. aeruginosa, followed by S. aureus⁴. S. aureus was isolated from 26/50 patients, from whom we obtained 51 isolates (Figure 1). Chromosomal analysis by PFGE yielded 34 patterns (pulsotypes) (Table 1). With the 34 patterns, we were able to identify nine clusters (designed as I-IX), which differed by approximately 85% in PFGE band similarity (Figure 1). The cluster with more members (thirteen) was cluster II. The results showed that nine patients (CF001, CF010, CF011, CF013, CF014, CF016, CF019, CF022, and CF029) were chronically infected (the same pulsotype was identified in two or more samples for two or more months) (Figure 2). Twelve patients were colonized with different pulsotypes: eleven with two different pulsotypes and one with three different pulsotypes (Table 1). Patients CF008 (colonized with Sau25) and CF014 (colonized with Sau06 and Sau14) died during the study period.

Figure 1 Dendrogram generated from PFGE analysis of 51 S. aureus isolates from pediatric CF patients. A representative PFGE profile of each pulsotype was used to construct the dendrogram. Each pulsotype frequency (number of isolates) is indicated, and each row indicates the patient in which the pulsotype was identified. Patients colonized by more than one pulsotype are indicated (*). A superscript indicates the frequency of each pulsotype identified per patient. The dotted line indicating 85% similarity was used to determine the cluster designation (I-IX). 

Figure 2 S. aureus pulsotypes were identified from 2016 to 2018. White boxes indicate a single pulsotype; colored boxes indicate a clone detected twice or more times. 

Antimicrobial susceptibility of S. aureus pulsotypes isolated from CF children

One of the 34 S. aureus pulsotypes identified was used to determine the antimicrobial susceptibility pattern to ten different classes of antibiotics: glycylcyclines, oxazolidinones, glycopeptides, folate pathway antagonists, fluoroquinolones, tetracyclines, macrolides, ansamycins, penicillins, lincosamides, and aminoglycosides (Table 2). We detected 3/34 (8.8%) MDR pulsotypes, and a low rate of resistance to vancomycin (8.8%), moderate resistance to tetracycline and erythromycin (20.6%), and clindamycin (17.7%). All the pulsotypes were susceptible to tigecycline, linezolid, and trimethoprim/sulfamethoxazole (Table 2). We found only two MRSA pulsotypes (5.9%): Sau08 and Sau16, both isolated from patient CF019. Sau16 was isolated three times at different times during sample collection (Figure 1 and Table 1). We identified both pulsotypes (Sau08 and Sau16) carrying the mecA gene by PCR (data not shown). These results also revealed that the two pulsotypes were MRSA with an MDR phenotype.

Pulsotypes of S. aureus induced cytotoxicity in human THP1-derived macrophages

The ability of S. aureus to induce cytotoxicity in human THP1 monocytes differentiated into macrophages (THP1m) was evaluated. The results showed that 30/34 (88.2%) pulsotypes were cytotoxic (> 10% of cytotoxicity) (Figure 3), 8/34 (23.5%) were highly cytotoxic (> 50% of cytotoxicity), and 14/34 (41.2%) were able to disrupt the THP1m monolayers (data not shown). To determine whether the cytotoxic effect was associated with pulsotypes carrying the PVL gene, we amplified the lukPV gene by PCR. The results showed that none of the analyzed pulsotypes carried the lukPV gene (data not shown).

Figure 3 Induction of cytotoxicity by S. aureus isolates in THP1macrophages: 5 x 10⁵ THP1 monocytes were differentiated into macrophages with 100 ng/mL of PMA for 24 h. Cells were infected with S. aureus isolates at MOI of 100 for 30 min. Once infected, cells were washed and incubated for 24 h. Supernatants were used to quantify macrophage cell death (cytotoxicity). Results were obtained from three independent experiments, each in duplicate (n = 6). Data were plotted as the mean ± SD and analyzed by one-way ANOVA and Dunnett’s multiple comparisons in relation to S. aureus -USA300. **p < 0.01; ***p < 0.001; ****p< 0.0001; ns: non-significant.