Background: Candida glabrata has emerged as an important nosocomial pathogen, yet little is known about its epidemiology.|

Aim: we aimed to analyze by polymorphic markers a collection of C. Glabrata isolates from hospital and non-hospital environment to look for inter-patients transmission at CHU Sfax hospital.

Methods: Two populations of C. Glabrata isolates were considered in our study. The first included 108 isolates from in-patients. The second population was collected from the out-patients and contain 37 isolates from vaginal and buccal sites.

Findings: A total of 145 Candida glabrata isolates were typed using three regular repeat units GLM4, GLM5 and MTI and irregular motif ERG3. Multilocus analyses resulted in the delineation of 62 genotypes and the identification of 8 new alleles and 33 new genotypes. Five genotypes G10, G42, G20, G13 and G22 were the most frequent and represented 51% of the entire population. Three genotypes G42, G10 and G20 predominated in both invasive and non-invasive population and in hospitalized and non-hospitalized. Moreover, the genotype G22 was absent in out-population and four patients infected with genotype G22 were hospitalized in the same intensive care unit and period.

Conclusion: We conclude that little variation exists between hospitalized and non hospitalized patients. The great differentiation was observed between buccal isolates and invasive population (2005- 2008). Besides, dissemination to multiple patients from a single source (health care worker, equipment) could be an important factor in the transmission of this yeast in our hospital.

Keywords: Candida glabrata ; Nosocomial Infection; Molecular Typing

During the last decade, incidence of invasive candidiasis with non-albicans Candida species has increased [1,2]. Candida glabrata is the most significant species that has emerged and now regularly ranks number two, after C. Albicans, as the etiologic agent of superficial and invasive candidiasis occurring in adults [3,4]. Reasons for this change in species distribution remain uncertain but may be partially due to the natural resistance of C. Glabrata to azole derivates, widely used since the 1980s [5,6]. Currently, despite increasing clinical concern, the epidemiology of C. Glabrata remains poorly known compared to that of C. Albicans. While C. Glabrata is considered a commensal of the human digestive tract, its natural reservoir is still uncertain.

The use of molecular technique to determine genotype of clinical isolates can provide useful clues for estimating the reproductive mode, geographic prevalence's, and cross transmission of bloodstream infections between patients [7- 12]. Possible transmission between patients has been suggested and clusters of invasive infections have been reported [13-15]. Molecular typing method would allow better understanding of the emergence of this species, notably in a hospital context. In the present study, we aimed to analyze by polymorphic markers a collection of C. Glabrata isolates from various sites in hospitalized and non hospitalized patients to look for association of variability with patient population.

A collection of C. Glabrata clinical isolates were collected in a prospective study between January 2005 and December 2011. The isolates were subdivided into outpatients and inpatients clinic populations of Habib Bourguiba Sfax-Hospital (Table 1). The first population included 108 isolates from in-patients and were subdivided as described in Table 1. Fifty nine invasive isolates from blood culture collected from different wards (intensive care unit, infectious disease unit, nephrology or endocrinology unit) of which 15 isolates were previously tested [3], and 49 isolates from urine samples. The second population contained 24 vaginal isolates and thirteen buccal isolates taken from oral cavities. All were collected from the outpatients at the same period [16]. All isolates were selected as only one isolate was considered for each patient and we were certain that patients from the outside hospital had never presented a positive culture from any site for C. glabrata and hadn't been hospitalized in the last two years.

For identification, isolates were plated onto Candiselect ID medium (Bio-rad, France) at 37°C for 48h. C. Glabrata was identified by its green color and this identification was confirmed by ID32C (Biomérieux, France) assimilation test and Glabrata RTT (Fumouze, Diagnostic, France). Fluconazole susceptibility testing was performed by measuring the Minimum Inhibitory Concentration (MIC) obtained through the use of the E-test (Biomérieux, France) as described by manufacture's protocol. A MIC ≥ 64 mg/L defined fluconazole resistance [17].

Genomic DNA was prepared from cell pellets obtained from 5 ml of fresh overnight culture in Yeast Peptone Glucose (YPG) medium. Genomic DNA was extracted using the nucleospin^™ Tissue method (Macherey-Nagel,Dueren, Germany) according to the manufacturer's instructions.

Molecular typing of invasive C. Glabrata isolates was performed as previously described by multiplex PCR using four microsatellite markers GLM4, GLM5, ERG3 and MTI [18,19]. Amplification was carried in 20 μl volume containing 2 μl of C. Glabrata DNA, 1X STR* buffer (Promega, USA), 0.2 mm deoxynucleoside triphosphate, 5 pmol each of the GLM4 and GLM5, 10 pmol of ERG3 primers, 20 pmol of the MTI primers and 1.25 U of Ampli Taq Gold DNA polymerase (Applied Biosystems, Courtaboeuf, France). After an initial step of 10 min at 95°C, the PCR included 30 cycles of 95°C for 30 s, 55°C for 30 s and 72°C for 1 min, followed by an additional step of 5 min at 72°C. One microliter of the PCR mixture was then added to 24 μl of formamide containing 0.5 μl of Genescan LIZ 500 marker (Applied Biosystems, Courtaboeuf, France) and denaturated for 2 min at 95°C. The PCR products were subjected to electrophoresis on an ABI 310 sequence analyzer and the data were analyzed with the Genescan software (Applied Biosystems, France).

The genetic differentiation between both populations was measured using Wright's Fixation Index (FST) computed with the Fstat V2.9.3 software developed by Goudet J. (http://www2. unil.ch/popgen/softwares/Fstat293dist.exe). FST is a measure of genetic divergence among populations [20]. The Fst value = 0.05 indicates little genetic divergence between populations while moderate genetic differentiation was considered as 0.05 < Fst < 0.15. A great genetic differentiation is defined as FST value ranged from 0.15 to 0.25.

Nei's genetic diversity was used as measure of gene diversity and differentiation among the populations [21]. To determine the extent of clonality and recombination in the different populations, the Index of Association (IA), rbard, rbars and Proportion of compatible pairs of loci were computed with Multilocus 1.3b software using 1000 randomizations tests to obtain reliable p-values and without considering repeated genotypes [22]. All test the null hypothesis of no linkage disequilibrium, which would indicate the absence of recombination. They are expected to be zero if populations are significantly freely recombining and greater than zero if there is association between alleles (clonality). The rbard statistic takes into consideration the number of loci tested and is considered a more robust measure of association and rbars test whether alleles tending in the same direction are positively or negatively associated. Prcompat should be 1 if all pairs of loci are compatible with strict clonality.

Phylogenetic relationships were estimated based on multilocus genotypes using the Neighbor-Joining method as implemented in http://pubmlst.org/. The effect of covariates on the cluster distribution was tested with the Chi-squared test.

In the present study, a total of 145 Candida glabrata isolates from inpatients and outpatients were typed using polymorphic markers (MTI, ERG3, GLM4 and GLM5). From hospitalized and non hospitalized patients, three and two separated subgroups according to site and time of isolation were characterized respectively (Table 2). Multilocus analyses resulted in the delineation of 62 genotypes (Table 2). Forty eight of the multilocus genotypes, corresponding to 33.1% of the total number of isolates, were represented by single isolate.

C. Glabrata genotyping were performed by only four polymorphic markers instead of six and allowed the identification of 8 new alleles (locus GLM4, alleles 246, 285 and 291; locus GLM5, alleles 307 and 268; locus ERG3, alleles 239 and 183; locus MTI, alleles 225) and 37 new genotypes (Table 2) in addition to 30 alleles and 37 genotypes previously described [14].

Overall, multilocus analysis of the 145 isolates showed that genotypes G10, G26, G9, G13 and G22 were the most frequent and represented respectively 18.6%, 14.48%, 7.58%, 5.51% and 4.8%. These multilocus genotypes together represented 51% of the entire population and were identified in all studied isolation sites except buccal site which has been characterized by unique genotypes (G41, G43, G51, G53-G60). These five genotypes represented 72.7% of the invasive candidiasis population (2005- 2008), 70.8% of the vaginal population (2005-2011), 45.9% of invasive candidiasis population (2009-2011) and urinary isolates (2005-2011) each.
When comparing isolates from inside and outside hospital, we noted that each population was dominated by the same genotypes and shared alleles and frequency in most cases. The frequency of some genotypes varied from one population to another although it was not significantly associated (p ≥ 0.05), i.e G22 was more common in invasive population (5.1%) and completely absent in outside of the hospital.

In order to assess differentiation among populations, Wright Fixation Index (Fst) was calculated between hospitalized and non hospitalized population (Fst = 0.0274). This result indicated a little genetic differentiation suggesting a non specific genetic structure of C. Glabrata population from inside or outside the hospital.

Wright fixation index was calculated for all pairwise combinations of the five sub-populations (Table 3). Invasive population (2005-2008) had a very similar structure to vaginal population and invasive population (2009-2011) was similar to urinary population with an Fst value 0.004 and 0.002, respectively. However, moderate genetic differentiation (0.05 < Fst < 0.15) was observed between vaginal isolates (2005-2011) vs invasive candidemia (2009-2011). A moderate genetic differentiation exist also between invasive candidemia (2005-2008), urine and vaginal isolates versus buccal samples (Table 3).

Genetic diversity varied from 0.85 for vaginal population to 0.95 in invasive isolates (2009-2011) and buccal samples (ID = 1) (Table 3). Buccal populations were considered the most divergent followed by invasive candidemia (2009-2011). In addition, there were a higher number of strains with a unique genotype within 13 among 13 isolates of buccal population than invasives isolates (2009-2011) (15 among 37 isolates), invasives isolates (2005-2008) (3 among 22) and vaginal isolates (2005- 2011) (5 among 24 isolates).

Neighbor-joining tree was constructed for the entire population based on the polymorphism of the four polymorphic markers as shown in Figure 1. There was no unique topology for intra or extra-hospital isolates. Isolates from different anatomic location were subdivised into different genotypes (Figure 1).

A neighbor-joining tree was constructed to show genetic relationship among 62 observed multilocus types in our combined population (Figure 2). Our population was subdivided into four major groups. The first group contains a unique multilocus type G30 from urinary population, the second contain G8 from vaginal population, the third contain the G10 which was the most frequent genotype and the fourth contains all other types including four of the major multilocus types G10, G26, G13 and G22.

To determine the extent of clonality and recombination in different populations, we used four measures of associations (IA, rbard, rbars and Proportion of compatible pairs of loci) [18]. For all studied sub-populations, a randomization test for linkage disequilibrium measures presumably rejected the null hypothesis of recombination as well as the total isolates considered to be a single population although proportion of compatible pairs of loci reveals some degree of recombination (prcompat ˂ 1) (Table 1).

The antifungal susceptibility to fluconazole of these strains was not different between hospitalized and non hospitalized patients (Table 1). Among 59 invasive C. Glabrata isolates and 86 non invasive isolates, only 15 isolates were resistant to fluconazole (10.34%). Genotypic analysis of these isolates revealed that they had different genotypes and were subdivided into 8 multilocus types (Figure 1) and the resistant phenotype was not significantly associated with a population or another (p ≥ 0.05).

We studied a genetic population structure of C. Glabrata isolates from inside and outside of the CHU hospital in Sfax (Tunisia). Polymorphic markers analysis revealed that all loci showed a high degree of diversity and notably each locus represented 6 -11 alleles (Table 2). Most of them were common to the majority of isolates but few were less common and were identified in only one or at least two isolates.

Genotypic distribution remains similar to that described in Tunisia in previous study when analyzing 85 unrelated isolates from different anatomical sites by six microsatellites [18]. In Tunisian population, the same genotypes (G10, G26 and G9) dominated. G10 was the most frequent type which was observed in 27 clinical isolates. The latter type was identified in phylogenetic tree as a single branch which seems to be stable and did not undergo natural evolution process. The overall abundance of G10 but also G26 and G9 suggest that at least some isolates with identical types may be clonally related by descent. In addition, we have also provided that there was a significant clonal component for the entire population (IA = 0.198).

In our report, we revealed that the inpatient population was not significantly differentiated from out-patient population (Fst < 0.05). Moreover, little differentiation was conducted for all pair wise population except buccal population which was considered moderately differentiated especially from invasive candidemia population. This latter result may be due to the specific virulence and resistance capacities of C. Glabrata in oral mucosa and digestive tract isolates [23,24].

Despite a lack of differentiation between the two populations (Fst = 0.0274), each of them included some genotypes which were not found in the other (Figure 1). Indeed, the inside hospital population was more divergent and showed the inclusion of 38 multilocus types. On the other hand, the outside hospital population was less divergent but showed the inclusion of 18 multilocus types not described in inside hospital.

In our setting, when analyzing 108 isolates from hospitalized patients and 37 from non hospitalized patients, we identified one multilocus type for 2.5 isolates for hospital environment and one multilocus type for 1.6 isolates in outpatients. Other studies, when analyzing intra-population diversity by Multilocus Sequence Type (MLST) established one multilocus type for every 4.8 patients in Baltimone in 2008 and one multilocus type for every 2.4 isolates in Atlanta from 1992 to 1993 [10,11,25]. These results demonstrated that our method may be more discriminative in examining population structure.

Most of C. glabrata fungemia are thought to originate from the endogenous flora of the host when host factors or environmental conditions allow them to switch from a commensal state to a pathologic state. In order to predominate as a pathogen, an isolate must first predominate as a commensal. In our study, three genotypes G10, G26 and G9 predominated in both invasive and non-invasive population and in hospitalized and non hospitalized patients. It means that invasive population was not differentiated from colonizing strains and indicated that any clone can be pathogenic in any infection site. This suggests also that endogenous flora of the host was the main source of infection and excludes the hypothesis that inter-patient transfer plays a role in the prevalence of at least most popular genotypes.

Moreover, some multilocus genotype G22 and G13 were more prevalent in invasive population. The genotype G22 were almost absent in out-population (vaginal and buccal samples). Our results suggested that some multilocus type may be stronger pathogen or that horizontal transfers can explain their highest prevalence inside the hospital. Furthermore, we notice that four patients (among 7 patients) infected with genotype G22 were hospitalized in 2009 in the same intensive care unit in the same period. This made that second hypothesis was more available and can be confirmed when analyzing a more large collection of isolates within and outside of the hospital.
It is also important to note that antifungal pressure exerted by some resistant genotypes can play a role in determining the population structure inside the hospital. It had been previously shown that there was no association between antifungal susceptibility and particular genotypes in previous study in Tunisia [3]. Similarly, Lott et al. [11] affirmed that fluconazole resistance profiles from invasive and non invasive population were not significantly different and were not associated with a particular sequence type. In our study, we didn't found an emergence of any of the major multilocus type and fluconazole resistance in inside or in the outside population hospital except for isolates collected in 2009 (4 isolates) in which two of them showed a resistant phenotype (Figure 1). Eleven percent of C. Glabrata isolates in this study were fluconazole resistant. Among them, 80% were isolated from hospitalized patients (p ≥ 0.05). Selection pressure exerted in patients inside the hospital perhaps due to antifungal treatment may be responsible.

We can conclude that very little differentiation exists between both hospitalized and non hospitalized patients, invasive and colonizing population and the dissemination to multiple patients from a single source (health care worker, equipment) could be an important factor in the transmission of this yeast in our hospital.

This study was financially supported by the Tunisian ministry of high education.No conflict of interest and approval states were not required. No specific funding was received for this study.

1. Pfaller MA, Diekema DJ. Epidemiology of invasive candidiasis: a persistent public health problem. Clin Microbiol Rev. 2007;20(1):133- 63.
2. Paluchowska P, Tokarczyk M, Bogusz B, Skiba I, Budak A. Molecular epidemiology of Candida albicans and Candida glabrata strains isolated from intensive care unit patients in Poland. Mem Inst Oswaldo Cruz. 2014;109(4):436-41.
3. Horn DL, Neofytos D, Anaissie E J, Fishman JA, Steinbach WJ., Olyaei AJ, et al. Epidemiology and outcomes of candidemia in 2019 patients: data from the prospective antifungal therapy alliance registry. Clin Infect Dis. 2009 ; 48(12):1695–1703. doi: 10.1086/599039.
4. Trick WE, Fridkin SK, Edwards JR, Hajjeh RA, Gaynes RP. Secular trend of hospital-acquired candidemia among intensive care unit patients in the United States during 1989–1999. Clin Infect Dis. 2002;35(5):627– 30.
5. Blot S, Janssens R, Claeys G, Hoste E, Buyle F, De Waele J J, et al. Effect of fluconazole consumption on long-term trends in candidal ecology. J Antimicrob Chemother. 2006;58:474–7.
6. Bassetti M, Righi E, Costa A, Fasce R, Molinari M P, Rosso R, et al. Epidemiological trends in nosocomial candidemia in intensive care. BMC Infect Dis.2006;6:21.
7. Abbes S, Sellami H, Sellami A, Makni F, Mahfoudh N, Makni H, et al. Microsatellite analysis and susceptibility to FCZ of Candida glabrata invasive isolates in Sfax Hospital, Tunisia. Med Mycol. 2011;9(1):10- 15. doi: 10.3109/13693786.2010.493561.
8. Pfaller MA, Messer SA, Boyken L, Tendolkar S, Hollis RJ, Diekema DJ. Variation in susceptibility of bloodstream isolates of Candida glabrata to fluconazole according to patient age and geographic location. J Clin Microbiol. 2003; 41(5):2176-9.
9. Pfaller MA, Messer SA, Hollis RJ, Boyken L, Tendolkar S, Kroeger J, et al. Variation in susceptibility of bloodstream isolates of Candida glabrata to fluconazole according to patient age and geographic location in the United States in 2001 to 2007. J Clin Microbiol.2009;47(10):3185-90. doi: 10.1128/JCM.00946-09.
10. Lott TJ, Fradeb PJ, Lockhart SR. Multilocus sequence type analysis reveals both clonality and recombination in populations of Candida glabrata bloodstream isolates from U.S. Surveillance Studies. Eukarotic cell. 2010;9(4): 619-25. doi: 10.1128/EC.00002-10.
11. Lott TJ, Frade JP, Lyon GM, Iqbal N, Lockhart SR.. Bloodstream and non-invasive isolates of Candida glabrata have a similar population structures and fluconazole susceptibilities. Medical Microbiology. 2012;50(2):136-42. doi: 10.3109/13693786.2011.592153.
12. Meeûs de T, Renaud F, Mouveroux E, Reynes J, Galeazzi G, Mallié M, et al. Genetic structure of candida glabrata populations in AIDS and non- AIDS patients. J Clin microbial. 2002;40(6):2199-206.
13. Asmundsdottir LR, Erlendsdóttir H, Haraldsson G, Guo H, Xu J, Gottfredsson M. Molecular epidemiology of candidemia: evidence of clusters of smoldering nosocomial Infections. Clin Infect Dis. 2008;47(2):17-24. doi: 10.1086/589298.
14. Vasquez JA, Dembry LM, Sanchez V, Vazquez MA, Sobel JD, Dmuchowski C, et al. Nosocomial Candida glabrata colonization: an epidemiologic study. J Clin Microbiol. 1998;36(2):421-6.
15. Nedretkoc A, kocagoz Koc NA, Erdem FS, Gunduz Z. Outbreak of nosocomial fungemia caused by Candida glabrata . Mycoses. 2002;45(11-12): 470-5.
16. Amouri I, Sellami H, Abbes S, Hadrich I, Mahfoudh N, Makni H, et al. Microsatellite analysis of Candida glabrata from recurrent vulvovaginal candidiasis. J Med Microbiol. 2012; 61(Pt 8):1091-6. doi: 10.1099/jmm.0.043992-0.
17. Mahmoud A. Ghannoum. M27-S4 :Reference method for broth dilution antifungal susceptibility testing of yeasts; fourth informational supplement. Wayne: Clinical and Laboratory Standards Institute;2012.
18. Abbes S, Sellami H, Sellami A, Hadrich I, Amouri I, Mahfoudh N, et al. Candida glabrata strain relatedness by new microsatellite markers. Eur J Clin Microbiol Infect Dis. 2012;31(1):83-91. doi: 10.1007/ s10096-011-1280-4.
19. Foulet F, Nicolas N, Eloy O, Botterel F, Gantier JC, Costa JM, et al. Microsatellite marker analysis as a typing system for Candida glabrata . J Clin Microbiol 2005; 43(9):4574-9.
20. Weir BS, Cockerham CC. Estimating F-statistics for the analysis of population structure. Evolution. 1984;38:1358-1370.
21. Nei M. Estimation of average heterozygocity and genetics distance from a small number of individuals. Genetics.1978; 89(3):583-90.
22. Agapow P-M, Burt A. Indices of multilocus linkage disequilibrium. Mol. Ecol. Notes. 2001;1(1-2):101-102. DOI: 10.1046/j.1471- 8278.2000.00014.x.
23. Enache-Angoulvant A, Bourget M, Brisse S, Stockman-Pannier C, Diancourt L, François N, et al. Multilocus microsatellite markers for molecular typing of Candida glabrata : application to analysis of genetic relationships between bloodstream and digestive system isolates. J Clin microbial. 2010;48(11):4028-34. doi: 10.1128/JCM.02140-09.
24. Li L, Redding S, Dongari-Bagtzoglou A. Candida glabrata : an emerging oral opportunistic pathogen. J Dent Res.2007;86(3):204-15.
25. Shin JH, Chae MJ, Song JW, Jung SI, Cho D, Kee SJ, et al. Changes in karyotype and azole susceptibility of sequential bloodstream isolates from patients with Candida glabrata candidemia. J Clin Microbiol. 2007;45(8):2385–91.