Prevalence and Risk Factors for Nasal and Oropharyngeal Carriage of Staphylococcus Aureus in Insulin-Dependent Diabetics Individuals

Individuals with insulin-dependent diabetes are a risk group for infections caused by Staphylococcus aureus. The objective of this study was to determine the prevalence and risk factors for nasal and oropharyngeal carriage of S. aureus and MRSA in insulin-dependent diabetic individuals from Botucatu, Additionally, susceptibility proling, detection of the mecA gene, SCCmec typing and molecular typing by PFGE and MLST were performed in nasal and oropharyngeal swabs of 312 subjects. The prevalence of S. aureus and MRSA was 30.4% and 4.8%, respectively. SCCmec type IV was the predominant type among isolates, although some carried SCCmec types I and II. MRSA and MSSA clones were detected among the isolates of this population. In addition, an important clonal lineage (ST398) was identied among resistant and susceptible isolates. Age and lung disease were negatively associated with S. aureus carriage, while lower-extremity ulcers were a risk factor for S. aureus carriage. For MRSA, only male gender was a risk factor. These data suggest widespread dissemination of MRSA in the insulin-dependent diabetic population studied, as well as the emergence of important lineages among these individuals.


Background
Diabetes mellitus is a progressive chronic disease characterized by high blood glucose levels, which is one of the most prevalent diseases in modern societies. Its treatment is often inadequate or absent [1]. It is estimated that more than 370 million people worldwide have diabetes and more than 5 million deaths were attributed to the disease and its complications in 2017 [2,3].
Diabetic individuals are known to be more susceptible to infections because of their elevated blood glucose levels and suppression of the immune response. In addition, neuropathy and reduced blood ow to the extremities are common in these individuals. Consequently, wounds tend to heal more slowly, increasing the risk of amputations and death [4,5].
Staphylococcus aureus is one of the leading causes of infections in diabetic individuals. Infections caused by methicillin-resistant S. aureus (MRSA) are associated with a higher mortality rate compared to infections caused by methicillin-susceptible S. aureus (MSSA) [2,6]. Studies suggest that nasal colonization with S. aureus and MRSA increases the risk of developing infections and that insulin use is a risk factor for colonization with MRSA in diabetic individuals [7]. However, little is known about the role of throat colonization in this population.
Considering their impact on the health of diabetic individuals, a better understanding of the epidemiology and risk factors attributed to colonization with S. aureus and MRSA is necessary. The objective of this study was to determine the prevalence and risk factors for nasal and oropharyngeal colonization with S. aureus and MRSA in insulin-dependent diabetic individuals. Additionally, antimicrobial susceptibility testing of the isolates, detection of the mecA gene, SCCmec typing, clonal pro ling, and molecular typing by MLST were performed.

Study design
This was a cross-sectional study conducted in the city of Botucatu, São Paulo, Brazil, whose estimated population is 146.497 inhabitants [8]. The size of the sample was calculated based on the number of insulin-dependent diabetic individuals registered at the municipal health department for receiving a glucose meter, which was 1,631 individuals. Using the calculation for proportions, we obtained a suggested n of 312 subjects. The subjects were selected randomly from the database of the health department. If possible, the subjects were recruited during home visits; however, many samples were collected during events promoted at Basic Health Units and at the headquarter of the Botucatu Association for Diabetes Support (ABAD in the Portuguese acronym). A questionnaire including the following data was applied to subjects who agreed to participate in the study: demographic data (gender and age); type of diabetes (1 or 2); time since diagnosis (years); time of insulin use (years); clinical data (comorbidities); presence of ulcers or amputations; tattoos; hospitalizations or medical procedures in the last year; infections and use of antimicrobials in the last year. All data were obtained by interview with the patient and/or legal representative following ethical standards.

Collection of microbiological specimens
Nasal and oral mucosa samples were collected from 312 insulin-dependent diabetic individuals living in the city of Botucatu, São Paulo, Brazil, using sterile swabs in transport medium. Samples from the anterior nares and oropharynx were obtained using one swab for each site. For the collection of nasal samples, the swab was immersed in 0.9% sterile saline and inserted into both nares, rotating it and gently pressing its end against the mucosa. The technique for oropharyngeal sampling consisted of immersion of the swab and passing it gently over the surface of the throat, avoiding contact of the examiner with the tongue.
The swabs were transported in Stuart medium to the Laboratory of Bacteriology, Department of Microbiology and Immunology, Institute of Biosciences, UNESP, and seeded onto plates containing Baird-Parker agar, a selective medium for Staphylococcus. After incubation for 48 h at 37 °C, the isolated microorganisms were identi ed.

Identi cation of Staphylococcus aureus
The microorganisms were submitted to Gram staining for observation of their morphology and speci c staining. After con rmation of these features, catalase and coagulase tube tests and biochemical tests (maltose, trehalose, and mannitol) were carried out for differentiating S. aureus from other coagulasepositive Staphylococci [9,10].
After DNA extraction with the Illustra Kit (GE Healthcare, Little Chalfont, Buckinghamshire, UK), the S. aureus isolates were con rmed genotypically by detection of the 16S rRNA gene [11] and the DNA SA442 fragment speci c for S. aureus [12].

Visualization of ampli ed products
The e ciency of the ampli cations was con rmed by electrophoresis on 2% agarose gel prepared in 0.5 M Tris-borate-EDTA (TBE) buffer. A 100-bp marker was used as molecular weight standard. The gel was stained with SYBR® Safe and photographed under UV transillumination.

Pulsed eld gel electrophoresis (PFGE)
The modi ed protocol of McDougal et al. [17] was used to type all S. aureus isolates by PFGE. The isolates were grown in BHI broth for 24 h. Next, 400 µL of the sample was added to a microtube and centrifuged at 12,000 rpm for 50 seconds. The supernatant was discarded and 300 µL TE solution (10 mM Tris, 1 mM EDTA, pH 8.0) was added. The samples were kept in a water bath for 10 min at 37 °C.
The samples were transferred to plug molds. The plugs were allowed to solidify and then placed in a 24well plate with 2 mL EC solution (6 mM Tris-HCl, 1M NaCl, 100 mM EDTA, 0.5% Brij-58, 0.2% sodium deoxycholate, 0.5% sodium lauroyl sarcosine) and incubated at 37 °C for at least 4 h. The EC solution was removed and the plugs were washed four times with 2 mL TE at room temperature at intervals of 30 min.
The SmaI enzyme (Fast Digest SmaI, Life Science, Canada) was used for restriction of genomic DNA. Electrophoresis was carried out in a CHEF-DR III System (BioRad Laboratories, USA) on 1% agarose gel prepared with 0.5 M TBE (Pulsed Field Certi ed Agarose, BioRad Laboratories, USA) under the following conditions: pulse switch time of 5 to 40 s for 21 h using a linear ramp; 6 V/cm; angle of 120°; 14 °C; 0.5M TBE as running buffer. The Lambda PFG Ladder (New England BioLabs) was used as molecular marker. The gel was stained with GelRed (10,000X in water;Biotium, USA) for 1 h and photographed under UV transillumination.
Similarity was analyzed with the BioNumerics 7.6 software (Applied Maths, Belgium). The dendrogram was constructed by the UPGMA method (Unweighted Pair Group Method with Arithmetic Mean), with band position tolerance and optimization adjusted to 1.2% and 1%, respectively.
Forty-ve isolates, including one MRSA, could not be typed with the SmaI enzyme and were therefore digested with the ApaI restriction enzyme.

Multilocus sequence typing (MLST)
MLST was performed according to the protocol of Enright et al. [18] by ampli cation and sequencing of seven housekeeping genes (arcC, aroE,glpF, gmk,pta, tpi, and yqiL). The PCR products were puri ed using the HiYield™ Gel/PCR Fragments Extraction Kit and the sequencing reaction is described below. The sequences were analyzed and compared to an online database (http://www.mlst.net) (2004). The sequencing reactions were performed in an ABI3500 8-capillary (50 cm) sequencer using POP7 as polymer (Applied Biosystems). The sequences were visualized (electropherogram) using the BioNumerics 7.6 program (Applied Maths, Belgium).

Statistical analysis
The Epi-Info for Windows software (version 7.26; ©Centers for Disease Control and Prevention, Atlanta, USA) was used for univariate analysis. Dichotomous variables were compared using nonparametric tests of proportion, X 2 test, and Fisher's exact test. Numerical variables were compared by the Mann-Whitney U test.
Multivariate analysis was performed with the SPSS 20.0 software (©SPSS, Inc.) using a logistic regression model. The outcomes of interest were the overall presence of S. aureus or the presence of MRSA irrespective of sampling site. The variables were selected using a backward stepwise strategy. The criterion for entry and permanence of the variables in the models was p < 0.1. Final statistical signi cance was set at p < 0.05.

Prevalence of Staphylococcus aureus and MRSA carriage
Ninety-ve of the 312 subjects included in the study were colonized with S. aureus; 15 were colonized with MRSA either in the nose or oropharynx. The overall prevalence of S. aureus was 30.4% (95%CI = 25.6%-35.8%) and that of MRSA was 4.8% (95%CI = 2.9%-7.8%). Regarding the sampling site, S. aureus was isolated exclusively from the nose in 44 (14.1%) individuals, exclusively from the oropharynx in 34

Determination of in vitro antimicrobial susceptibility
All 112 isolates identi ed as S. aureus were submitted to in vitro antimicrobial susceptibility testing as previously described. The disc diffusion method revealed 11 strains that were resistant to oxacillin and cefoxitin, including seven resistants to both drugs and four that were resistant only to cefoxitin (inhibition zone ≤ 21). It should be noted that four of the isolates that were susceptible to both drugs exhibited resistance by the genotypic method.
Among all isolates, only one was resistant to sulfamethoxazole/trimethoprim and there was no case of resistance to quinupristin/dalfopristin or linezolid. In addition, the MIC50 and MIC90 for vancomycin were 0.50 and 1.0 µg/mL, respectively, and all isolates were susceptible (Supplementary Material Table S1).

Detection of the mecA gene and characterization of SCCmec
Fifteen of the 112 S. aureus isolates carried the mecA gene of methicillin resistance. There was a predominance of the SCCmec type IV among isolates (n = 10), but three isolates harboring SCCmec type I and two harboring SCCmec type II were also identi ed. Most individuals harboring MRSA isolates reported no contact with the hospital environment in the last year, except for one subject with SCCmec type II and one with SCCmec type IV.

Risk factors for S. aureus and MRSA carriage
The results of univariate and multivariate (logistic regression model) analysis to identify risk factors for S. aureus and MRSA carriage are shown in Table 1 and 2. With respect to the study of risk factors for S. aureus carriage, univariate analysis showed a negative association with age (median 61years, p = 0.04), heart disease (p = 0.03) and lung disease (p = 0.02), and a positive association only with the presence of lower-extremity ulcers (p = 0.02). However, the only factors that remained independently associated with S. aureus carriage in insulin-dependent diabetic individuals were age (OR = 0.98, 95% CI = 0.93-0.99, p = 0.02) and lung disease (OR = 0.31, 95% CI = 0.10-0.92, p = 0.03), which were negatively associated. A positive association was only observed with lower-extremity ulcers (OR = 2.44, 95% CI = 1.11-5.34, p = 0.03).  Note. All values are reported as number (%) unless otherwise speci ed. Signi cant associations are indicated in bold. OR = odds ratio; CI = con dence interval; CNS = central nervous system.

Determination of the clonal pro le of S. aureus and MRSA by pulsed eld gel electrophoresis (PFGE)
A total of 112 S. aureus isolates were analyzed by PFGE. Forty-ve of them could not be typed repeatedly with SmaI, including one MRSA isolate. However, molecular typing of all 45 isolates was possible using the ApaI enzyme.
For clonal pro le analysis, one dendrogram was constructed for susceptible S. aureus isolates (MSSA) and one for resistant isolates (MRSA) using SmaI and ApaI, which permitted to identify clusters with similarity ≥ 80% in both groups.  555N, 659N, and 665N). All of them were isolated from different individuals and harbored SCCmec type IV. Cluster B contained three strains, two of them showing 100% similarity (72O and 615N). The MRSA isolate typed with ApaI is shown in Fig. 2 (b). None of the isolates grouped with international clones.
Note a. Isolates showing > 80% similarity (clusters A and B) after digestion with SmaI. b. Band pattern of strain 735N obtained by digestion with ApaI. N = nasal mucosa; O = oropharyngeal mucosa; S = susceptible; R = resistant. * International clones used as controls.  Figure S1). Cluster A contained two isolates from the nasal and oropharyngeal mucosa of the same subject, demonstrating colonization of different sites with the same isolate. Similarly, in cluster B, two isolates from the nasal and oropharyngeal mucosa of the same subject were grouped with one nasal isolate of another subject. On the other hand, the four isolates of cluster C were obtained from unrelated individuals and in cluster D all isolates were from the oral mucosa of different subjects. These ndings suggest widespread dissemination of S. aureus in the community.
Four (26.7%) of the 17 subjects colonized at both sampling sites (nasal/oropharyngeal) carried the same isolate in the nasal and oral mucosa, while ve (33.4%) carried different S. aureus strains in the nose and throat. Interestingly, two patients concomitantly colonized with S. aureus in the nose and throat had their strains (nasal and oropharyngeal) typed after digestion with different restriction enzymes. Isolates 691O and 747N were typed with SmaI, while 691N and 747O could only be typed with ApaI, con rming that they are different S. aureus strains.
Analysis of typable isolates with ApaI revealed three major clusters (A, C, and D) and three minor clusters of 3 isolates (B, E, and F). In cluster A which contained 14 isolates with 85.6% similarity, two isolates were from the nasal and oropharyngeal mucosa of the same patient and the remaining from different subjects. The same was observed for clusters C and D. In cluster D, the isolates were grouped with a strain previously identi ed in another study as ST398 (strain 76N). These ndings are shown in Supplementary Material Figure S2.

Molecular typing of S. aureus and MRSA by multilocus sequence typing (MLST)
Based on the clusters obtained by PFGE, nine S. aureus isolates (four MSSA and ve MRSA) were selected for molecular typing by the MLST technique. Typing of the MRSA isolates revealed a predominance of sequence type ST5 in three of the four isolates analyzed and one isolate with ST8. In addition, the strain typed with ApaI by PFGE was characterized as ST398 (Fig. 3)  Among the clusters obtained by PFGE of the MSSA isolates that could not be typed with SmaI and were digested with ApaI, four lineages were selected for MLST. There was a predominance of ST398 (n = 3). However, one isolate exhibited divergence in the allele of the arcC gene and was sent to the curator of the MLST database (http://www.mlst.net) for identi cation of the ST. This isolate was identi ed with ST 6133 (Fig. 4).

Note
Clustering of isolates digested with ApaI that were analyzed by MLST. All isolates except for 735N were susceptible to methicillin (MSSA

Discussion
Bacterial resistance has become a great global threat. Staphylococcus aureus plays a particularly important role in this scenario by causing infections that range from super cial to severe and potentially fatal systemic infections, in addition to its ability to acquire resistance to multiple drugs. In the past, the pathogen was mainly found in hospital environments, but we are now witnessing an increase particularly in resistant isolates (MRSA) acquired in the community that are genetically different from traditional nosocomial strains [19][20][21]. Furthermore, studies suggest that staphylococcal infections are preceded by the process of colonization [2,[22][23][24][25][26] and at least one third of colonized healthy adults are at risk of developing subsequent invasive infections [19,20].
Diabetic individuals are considered a risk group for skin infections such as those caused by S. aureus and are also more prone to developing severe systemic infections [26]. Studies have shown that the prevalence of nasal colonization with S. aureus (27%-56.6%) and MRSA (1%-7.3%) varies according to geographic location and that the use of insulin is a risk factor for nasal MRSA carriage in diabetic individuals [7].
The prevalence of colonization of S. aureus and MRSA in insulin-dependent diabetic individuals was 30.4% and 4.8%, respectively. Our results differ from those reported by Kutlu et al. [27], who analyzed 304 patients and found a prevalence of 41.9% for S. aureus and of 9.9% for MRSA among diabetic individuals. Alizargar et al. [7] also obtained a higher prevalence of S. aureus (42.5%) and MRSA (24.7%). In a population-based survey conducted in the same city, Pires et al. [29] found a similar prevalence among community-dwelling individuals, with an overall prevalence of S. aureus of 32.7%. However, the prevalence of MRSA was higher in our study (4.6% vs 0.9%). In a recent study, Lin et al. [2] found a lower prevalence of S. aureus and MRSA than that obtained in the present study (16.4% of S. aureus and 2.8% MRSA). The authors suggested that the presence of different microorganisms in the microbiota of these individuals causes competition for the same site, which could explain the variation found in prevalence studies. None of the studies included oral mucosa as a potential site of colonization.
It should be noted that 10.9% (n = 34) of the subjects were colonized exclusively in the oral mucosa, six of them with MRSA. This nding reinforces what was reported of Partida et al. [29] that colonization of the oral mucosa can compromise control measures of pathogen dissemination since the throat is not part of routine screening.
The MRSA isolates were identi ed by phenotypic methods (disc diffusion) and by PCR for detection of the mecA gene. The latter is considered the gold standard for detecting methicillin-resistant isolates. The CLSI [13] recommends the use of the cefoxitin disc for the detection of methicillin resistance because of its greater sensitivity in identifying isolates carrying the mecA gene. The results of the present study con rm the greater sensitivity of the cefoxitin disc compared to oxacillin, with only seven of the 15 isolates being identi ed as resistant when the oxacillin disc was used, while 11 were resistant to cefoxitin. However, we found four isolates carrying the mecA gene that did not exhibit phenotypic resistance to cefoxitin or oxacillin.
The most common agent isolated from infections in diabetic individuals is S. aureus, with a high prevalence of MRSA [30]. In a recent study, Chaudhry et al. [6] investigated clinical isolates and observed that 84% of the S. aureus isolates were MRSA and 20% were resistant to vancomycin. In our study, the prevalence of MRSA was lower and none of the isolates was resistant to vancomycin, although one MRSA isolate had a MIC of 1.5 µg/mL, indicating a potential therapeutic risk [31][32][33]. Two other MSSA isolates had a vancomycin MIC of 1.5 µg/mL.
In the present study, the analysis of risk factors revealed a positive association only with lower-extremity ulcers, which is consistent with literature ndings showing that the same isolate colonizing the nares was found in foot ulcers and wounds [34,35]. Age was negatively associated with S. aureus colonization, with an odds ratio of 0.98, indicating a reduction of 2% in the risk of colonization for each additional year of age. Similar data have been reported by Pereira-Franchi et al. [36] Lung disease was also negatively associated with S. aureus colonization, a fact that might be related to colonization of the respiratory tract by other microorganisms that are competing with S. aureus. Mueller et al. [37] also found a negative association with age but, in contrast to our ndings, lung disease was a risk factor for S. aureus colonization.
Multivariate analysis identi ed male gender as a risk factor for colonization with MRSA. These results are in contrast with those reported by Pereira-Franchi et al. [36] who found no risk factors for MRSA colonization.
With respect to clonality of the isolates, one important MRSA clone was detected in the community insulin-dependent diabetic individuals studied. Five of the 15 diabetic individuals with MRSA carried isolates that belonged to the same cluster (A), all of them harboring SCCmec type IV. It is important to emphasize that isolates 639O and 659N present in this cluster harbor ST5, possibly re ecting a common origin. There was also a smaller cluster (B) with three isolates obtained from three different subjects, one of nasal origin (615N-ST5-SCCmec I) and two of oral origin. One of them (72O) harbored ST8 and SCCmec type IV. This fact reinforces the importance of throat colonization, which could be a route of transmission within the population studied. Other studies involving individuals from the same city and region also found MRSA ST5-SCCmec type IV and ST8-SCCmec type IV, suggesting that these strains are prevalent in the region [28,[38][39][40][41][42][43].
Studies suggest a high clonal diversity among S. aureus isolates, particularly among MSSA [44]. We observed the formation of only four clusters among the MSSA isolates, but several isolates from the same individual were grouped together, with the identi cation of ve individuals with different S. aureus in the nose and throat. There was a large number of isolates that showed similarity with the ST398 clonal lineage. Isolates of this lineage are not digested with SmaI because of a yet unknown restriction/DNA methylation system in the genus Staphylococcus spp. However, these strains can be digested with the enzyme ApaI [45, 46]. Forty-ve isolates that could not be typed by conventional PFGE (restriction with SmaI) were identi ed, which formed three major clusters and four clusters with a smaller number of strains after digestion with the restriction enzyme ApaI. Among the 45 isolates, four were typed by MLST and were characterized as ST398. This fact was also observed by Souza et al [42].
The ST398 clonal lineage has been associated with infection and colonization of humans and domestic animals, such as dogs, horses and pigs, in many countries around the world [21]. This lineage is called livestock-associated S. aureus and was described for the rst time among both MSSA and MRSA on pig farms in France [45,46]. Since then, ST398 has spread rapidly to other animals and has been increasingly related to infections not only in rural workers but also in people and animals without risk factors [13,47]. Although susceptible to oxacillin, this S. aureus lineage is associated with severe infections, as reported by Bonesso et al. [48] in patients with ventilator-associated pneumonia in whom the infection was fatal in most cases.
Our nding demonstrated a predominance of SCCmec type IV among isolates, in agreement with the ndings of other prevalence studies on non-diabetic individuals conducted in the State of São Paulo [38,42]. However, SCCmec types I and II were also detected, which are commonly found circulating in health services. The presence of SCCmec type II has also been reported by Pereira-Franchi et al. [36].
The present study provides important data about the epidemiology of S. aureus and MRSA in a population of insulin-dependent diabetic individuals. The prevalence found is lower than that reported in the international literature, probably because the individuals were living in the community, most of them rarely visiting the hospital, and were generally in good health. A higher prevalence of MRSA colonization was observed in the population studied. Within this context, screening for oral colonization is extremely important since some individuals were colonized only at this body site.
The isolates analyzed had a low rate of resistance to the tested drugs, with only one isolate being resistant to sulfamethoxazole-trimethoprim; however, the prevalence of MRSA was higher than that found in a population-based study conducted in the same city on healthy individuals [16]. In the population studied here, clones were detected among the MSSA and MRSA isolates and an important clonal lineage (ST398) was identi ed. These data suggest widespread dissemination of MRSA in the insulin-dependent diabetic population studied, as well as the emergence of important S. aureus lineages in these individuals.

Ethical approval and consent to participate
The study was approved by the Ethics Committee of the Botucatu Medical School, Brazil (Approval No. 1.107.685). All data were obtained by interview with the patient and/or legal representative following ethical standards, by signing the consent form.

Consent for publication
Not Applicable.

Availability of data and material
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The author(s) declare no competing interests.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.