mycoses
Diagnosis,Therapy and Prophylaxis of Fungal Diseases
Review article
A review of molecular techniques to type Candida glabrata isolates
S. Abbes, I. Amouri, H. Sellami, A. Sellami, F. Makni and A. Ayadi
Laboratoire de biologie moléculaire parasitaire et fongique, faculté de médecine  Rue magida Boulila, Sfax, Tunisia
Summary Candida glabrata has emerged as a common cause of fungal infection causing mucosal
and systemic infections. This yeast is of concern because of its reduced antifungal
susceptibility to azole antifungals such as fluconazole. A clear understanding of the
epidemiology of Candida infection and colonisation required a reliable typing system for
the evaluation of strain relatedness. In this study, we discuss the different molecular
approaches for typing C. glabrata isolates. Recent advances in the use of molecular
biology-based techniques have enabled investigators to develop typing systems with
greater sensitivities. Several molecular genotypic approaches have been developed for
fast and accurate identification of C. glabrata in vitro. These techniques have been
widely used to study diverse aspects such as nosocomial transmission. Molecular
typing of C. glabrata could also provide information on strain variation, such as
microvariation and microevolution.
Key words: Candida glabrata, typing, polymorphism.
DNA probes,10,11 multilocus sequence typing
Introduction
(MLST)6,11 and more recently microsatellite analysis of
Candida species are the most common opportunistic markers.12,13 In this study, we discuss the molecular
fungal pathogens in humans, with Candida albicans tools for genotyping analyses of C. glabrata isolates to
being the most prevalent pathogen in mucosal and understand the epidemiology of this pathogenic yeast.
systemic fungal infections.1 In addition to C. albicans,
Candida glabrata is now emerging as an important agent
Restriction enzymatic electrophoresis
in both mucosal and bloodstream infections. The
prevalence of C. glabrata has increased in the last decade Restriction enzymatic electrophoresis (REA) is one of the
and this species now ranks as the second or third most first applied techniques for identification and typing
frequently isolated Candida species from all reported pathogenic fungus. REA was initially used by Scherer
cases of candidiasis.2,3 To investigate the epidemiology and Stevens [14], who predicted that this method,
of this pathogen, it is imperative to develop DNA because of its stability and reproducibility, would be
fingerprinting methods which can assess genetic dis- useful for epidemiological studies of Candida albicans.14 16
tance between independent isolates in broad epidemio- This method is straightforward. DNA is extracted from
logical studies. Different methods have previously been spheroplasts, digested with one or more endonucleases
applied to differentiate C. glabrata isolates, including and separated by electrophoresis in an agarose gel. The
random amplification of polymorphic DNA,4,5 pulsed- banding pattern of digested DNA is then visualised by
field gel electrophoresis (PFGE),6,7 multilocus enzyme staining with ethidium bromide. Separation depends
electrophoresis (MLEE),8,9 fingerprinting with complex upon the percentage of agarose in the gel, the
electrophoresis time, the voltage and the particular
Correspondence: A. Ayadi, Laboratoire de biologie moléculaire parasitaire
endonuclease(s) employed. The pattern is based on
et fongique, faculté de médecine, Sfax 3029, Tunisia.
different fragment lengths determined by the restriction
Tel.: [phone]. Fax: [phone]
sites identified by the particular endonuclease(s) used.
E-mail: [e-mail]
Variations among strains can occur as a result of changes
Accepted for publication 25 May 2009 in restriction site sequences, such as deletions and
Ó 2009 Blackwell Verlag GmbH " Mycoses 53, 463 467 doi:10.1111/j.1439-0507.2009.01753.x
S. Abbes et al.
insertions in the sequences between recognition sites. methods, particularly pulsed gel electrophoresis.18 How-
REA was widely investigated to know the epidemiology of ever, the limit of this methodology was the low-stringency
nosocomial infection. Vasquez et al. [17] typed isolates conditions which induce a poor reproducibility of typing
from 24 patients admitted to medical intensive care unit results.24
and bone marrow transplant unit, the typing of three
environmental isolates showed common restriction
Pulsed-field gel electrophoresis
profile to five patients. Arif et al. [18] assigned that REA
with HinfI was a more suitable method than other Pulsed-field gel electrophoresis was based on variation
genotyping systems. Similarly, Vasquez et al. [19] exam- in electric field vectors that were able to separate large
ined various yeast with different molecular methods, and molecular weight DNA. PFGE has been successfully
suggested that PFGE was more sensitive than REA for applied as a tool for typing Candida strains19,27,28 and
C. lusitaniae, C. parapsilosis, C. tropicalis and for C. glabrata. was used in both identification and typing of C. glabrata
It was also suggested that REA was more difficult than strains.29,30 This methodology proved to be a reliable
PFGE for strain differentiation. So, the lack of sensitivity and reproducible method for C. albicans and C. glabrata,
of strain typing and the difficulty of reading complex and is more sensitive than DNA enzymatic restriction
pattern made difficult to delineate the epidemiology of C. for typing C. albicans isolates.19,31,32 Two electropho-
glabrata.19 Recently, genotyping systems based on retic karyotyping systems exist [e.g. contour-clamped
restriction enzymatic analysis were less used for typing homogeneous electric field gel electrophoresis (CHEF)
C. glabrata isolates,20 but REA may be useful for the and transverse alternating field electrophoresis (TAFE)].
identification and strain delineation of common and CHEF seems to be more useful than TAFE for C. glabrata
emerging Candida species.21 typing system. Using PFGE typing system, several
authors demonstrated that the same C. glabrata strain
remained colonised in the patients for over several
Random amplification of DNA
monthsÕ time and that resistant and susceptible isolates
Random amplification of DNA (RAPD), described by remained the same DNA type.6,33 While leading a
Williams et al. [22], is one of the most applied techniques comparative study between PFGE and MLST for 80
for the survey of the epidemiology of Candida glabrata and superficial C. glabrata isolates collected from 27 inten-
other fungi.23,24 Using random primers of approximately sive care units, Lin et al. [6] suggested that PFGE (DI =
10 bases, amplicons throughout the genome are targeted 0.99) exhibited more discriminatory power than MLST
and amplified. Amplified products are separated on (DI = 0.85). PFGE could efficiently divide the groups
agarose gel and stained with ethidium bromide. The defined by MLST. PFGE was also able to show fine
genetic variation analysis based on RAPD allows proper variations of a single strain. Several authors analysed
genetic diversity due to its capacity to generate random genotypic variability and azoles susceptibility among
markers from the entire genome. Contradictory findings sequential bloodstream isolates by electrophoretic
have been found by some authors.4,10 Boldo et al. [4], karyotyping showed minor karyotypic changes or
when analysing 47 clinical isolates from several geo- microevolution.33,34 The results of PFGE have shown
graphical origins and diverse body sites, suggested that a high rate of inter-laboratory agreement24 and it
the data obtained by five RAPD markers showed no continues to be a reliable technique for C. glabrata
differentiation among strains (average heterozygosity, infection. So, the major disadvantages of PFGE are the
H0 = 0.372). On the other hand, Lockhart et al. [10] initial investment cost of the equipment, the inability to
selected nine RAPD markers from 30 primers that were run more than 20 samples at one time and the amount
able to differentiate among 39 isolates from different of time required for the plug preparation and for
geographical populations. The genetic diversity obtained electrophoresis.19
for C. glabrata was low4,25 compared with that reported
by other authors for the diploid species C. albicans.9,26
Southern blot hybridisation
However, the RAPD data for C. tropicalis and C. guillier-
mondi showed a discrimination power similar to those To visualise particular fragments in the pattern, we
obtained for C. glabrata. For this reason, it was not can probe restriction profile with radiolabelled or
possible to associate the low genetic diversity with the biotinylated DNA sequences that recognise one or more
mode of reproduction. fragments as a result of sequence homology. Using a
This technique is methodologically easier, less time screen for complex genomic fragments containing
consuming and cheaper than the older genomic typing moderately repetitive sequences, Lockhart selected two
464 Ó 2009 Blackwell Verlag GmbH " Mycoses 53, 463 467
Molecular methods for typing C. glabrata isolates
probes Cg6 and Cg12 which generate complex South- reason for which this technique has not been adopted by
ern blot hybridisation patterns with EcoRI-digested the majority of clinicians in addition to the relatively
C. glabrata DNA. The capacity and specificity of the time consuming to combine the data from at least 10 or
probes to measure genetic distance between indepen- more enzymes that provide variability among isolates.20
dent isolates were verified.10 They found a strong
geographical localisation of C. glabrata strains both
Multilocus sequence typing
between continents and between cities within a conti-
nent.10 They demonstrated that both Cg6 and Cg12 To obtain a higher resolution DNA fingerprinting
probes discriminate microevolution within sequential system based on direct sequence comparison, a MLST
isolates of C. glabrata,10 as observed in clonal population system was developed by Bougnoux et al.[37] to type
of C. albicans grown over many generations with C. albicans isolates. The data from alleles at multiple loci
Ca3 probes.35,36 The probes also contain invariant are combined to access genetic relatedness. The study of
sequences which facilitate normalisation in computer- genetic variation by MLST based on differences in DNA
assisted analysis that can be used in large epidemiolo- sequences is more sensitive than that accessed by
gical studies. When analysing diverse geographical differences in protein mobility.
collection of 107 clinical isolates, Dodgson et al. [11] Dodgson et al. [11] were the first to develop
demonstrated that Cg6 and Cg12, in contrast to RAPD C. glabrata MLST. They used MLST for two main
and MLST, discriminate between all isolates in all objectives. Firstly, the typing was for geographical
groups and were the better methods when analysing repartition of isolates. Analysing three geographically
microevolution or nosocomial transmission. The major diverse collections of clinical isolates, they identified five
disadvantage of Southern blot was the radioactive probe major clades, three of them exhibited significant
and the complexity of binding patterns needing auto- geographical bias.11 Secondly, MLST was performed to
matically identification by computer assisted systems.20 test the evidence of genetic recombination. Thirty-four
sequence types were defined and fourteen examples of
phylogenetic incompatibility were found. Thus, they
Multilocus enzyme electrophoresis
conclude that although C. glabrata has a predominantly
The MLEE analysis is based on allelic frequency that clonal population structure, the multiple phylogenetic
assesses isozymes or allozymes. MLEE can discriminate incompatibilities suggested that recombination occurred
among the gene products of different alleles for a during the evolution of C. glabrata, and may infre-
number of loci. Meeus et al. [8] have genotyped 63 quently occur.38
C. glabrata isolates for 33 putative gene enzymatic loci. Recently, Lin et al. [6] analysed 25 patients with
This method was verified for C. albicans by a cluster multiple isolates assume that MLST (DI = 0.85)
analysis of a set of test isolates in which MLEE, RAPD exhibited less discriminatory power than PFGE with
and Southern blot hybridisation with fingerprinting BssHII.
probes were compared and parity was demonstrated.9 This system is highly effective to examine population
However, for C. glabrata, a low genetic differentiation structure, but is not suited for studies of nosocomial
between distant hospitals (Montpellier and Paris) was infection or microevolution. The C. glabrata MLST
found and no correlation was noted with anatomic groups revealed a lower variable-site percentage of
origins or human immunodeficiency virus-positive or 1.6%, which is much less than that of C. albicans,39,40
negative (HIV+, HIV)) patients.4,8 Boldo et al. [4] and then a less discriminatory power was achieved. This
analysed 47 clinical isolates from several geographical may be due to the haploid nature of C. glabrata.
origins and diverse body sites, and estimated for 11
enzymatic loci a high level of genetic relatedness
Microsatellite markers
among isolates and revealed no genetic differentiation
(H0 = 0.055) among them. No association with their Genotyping isolates with microsatellite markers has
geographical origin and clinical characteristics was been recently described for C. glabrata.12,13 The poly-
found. Relationship between resistance to fluconazole morphism of microsatellites was evaluated by PCR using
and particular genotype was also searched but no fluorescent primers and an automatic sequencer as
correlation was found.8 However it can happen that the already reported for C. albicans. Microsatellites represent
percentage of mutation is so high that it is in association another class of genotyping defined as short tandem
with several MLEE combinations.4 MLEE does not show repeats of two to six nucleotides known to be highly
a significant differentiation among C. glabrata isolates, polymorphic.
Ó 2009 Blackwell Verlag GmbH " Mycoses 53, 463 467 465
S. Abbes et al.
Foulet et al. [12] adopted three polymorphic micro- References
satellite markers, RPM2, MTI and ERG3, for rapid
1 Safdar A, Chaturvedi V, Koll BS, Larone DH, Perlin DS,
typing of 138 isolates of C. glabrata and found a
Armstrong D. Prospective, multicenter surveillance study
discriminatory power equal to 0.84. Irregular distribu-
of Candida glabrata: fluconazole and itraconazole suscepti-
tion of C. glabrata population was observed which was
bility profiles in bloodstream, invasive and colonizing
similar to that observed with MLST. Twenty-four per
strains and differences between isolates from Three Urban
cent of the isolates belonged to the same sequence
teaching Hospitals in New York City. Antimicrob Agents
type.40 Recently, a multiple-locus variable-number Chemother 2002; 46: 3268 72.
2 Richardson MD. Changing patterns and trends in systemic
tandem-repeat analysis (MLVA) using six microsatellite
fungal infections. Antimicrob Agents Chemother 2005; 56:
markers was assessed in 127 C. glabrata isolates. Thirty-
5 11.
seven different genotypes, stable both in vitro and in vivo,
3 Malani A, Hmoud J, Chui L, Carver PL, Bielaczyc A,
were observed leading to discriminatory power (DI =
Kauffman CA. Candida glabrata fungemia: experience in
0.902).13 It seems that an ecological advantage for
a tertiary care center. J Clin Microbiol 2005; 41: 975
some genotypes exists.12
81.
Microsatellite marker system for DNA fingerprinting
4 Boldo XM, Villa-Tanaca L, Zuniga G, Hernandez-Rodri-
was easy to assay, adaptable to large series and can be
guez C. Genetic diversity among clinical isolates of Candida
used as a typing system to investigate clinical issues
glabrata analysed by randomly amplified polymorphic
such as nosocomial transmission of isolates or the origin
DNA and multilocus enzyme electrophoresis analyses.
of infective strains.12,13 However, the necessity to use J Clin Microbiol 2003; 41: 4799 804.
5 Song X, Sun J, Store G, Eribe ERK, Hansen BF, Olsen I.
fluorescently labelled primers with elevated cost and an
Genotypic relatedness of yeast in thrush and denture
automatic sequencer limits the utilisation of this
stomatis. Oral Microbiol Immunol 2006; 21: 301 8.
technology.
6 Lin CY, Chen YC, Lo HJ, Chen KW, Li SY. Assessment of
Candida glabrata strain relatedness by pulsed field gel
Conclusion electrophoresis and multilocus sequence typing. J Clin
Microbiol 2007; 45: 2452 9.
Following the widespread use of antifungal therapy
7 Redding SW, Kirkpatrick WR, Saville S, Coco BJ, White W,
and the increased number of immunocompromised
Fothergill A. Multiple patterns of resistance to fluconazole
patients, C. glabrata has emerged as an important
in Candida glabrata isolates from a patient with oropha-
opportunistic pathogen in superficial and invasive
ryngeal candidiasis receiving head and neck radiation.
infections. The typing of C. glabrata isolates has been J Clin Microbiol 2003; 41: 619 22.
8 de Meeus T, Renaud F, Mouveroux E, Reynes J, Galeazzi G,
developed mainly to detect the origin of nosocomial
Mallié M. Genetic structure of Candida glabrata populations
transmission.17,41,42 Then subsequently several objec-
in AIDS and non-AIDS patients. J Clin Microbiol 2002;
tives have been added, such as analyse variability
40: 2199 206.
between geographically distant isolates; recently, char-
9 Pujol C, Joly S, Shawn R et al. Parity among the
acterisation of resistant strains and determination if
randomly amplified polymorphic DNA method, multilo-
resistance is the result of a selection or an induction of
cus enzyme electrophoresis, and Southern blot hybrid-
initially sensitive strains. Some genotyping systems
ization with the moderately repetitive DNA probe Ca3
such as RFLP (due to the difficulties in interpretation)
for fingerprinting Candida albicans. J Clin Microbiol 1997;
and RAPD (due to the non-reproducibility) are not
35: 2348 58.
used anymore. On the other hand, fine variations
10 Lockhart SR, Joly S, Pujol C, Sobel JD, Pfaller MA, Soll DR.
between strains have been searched by a polymor- Development and verification of fingerprinting probes for
Candida glabrata. Microbiology 1997; 143: 3733 46.
phism of lengths of microsatellite sequences. The
11 Dodgson AR, Pujol C, Denning DW, Soll DR, Fox AJ.
principal reason for typing was to show that the
Multilocus sequence typing of Candida glabrata reveals
population of C. glabrata is mainly clonal and that a
geographically enriched clades. J Clin Microbiol 2003;
particular profile for the resistant strains does not exist.
41: 5709 17.
Southern blot with Cg6 and Cg12, karyotyping anal-
12 Foulet F, Nicolas N, Eloy O et al. Microsatellite Marker
yses and microsatellite marker analysis were the most
analysis as a typing system for Candida glabrata. J Clin
adequate techniques for typing C. glabrata strains.
Microbiol 2005; 43: 4574 9.
However, PFGE is not easy to achieve; therefore, it
13 Grenouillet F, Millon L, Bart JM et al. Multiple-locus
was not chosen for epidemiological issues such as
variable-number tandem-repeat analysis for rapid
nosocomial transmission or the origin of infective
typing of Candida glabrata. J Clin Microbiol 2007;
strains. 45: 3781 4.
466 Ó 2009 Blackwell Verlag GmbH " Mycoses 53, 463 467
Molecular methods for typing C. glabrata isolates
14 Scherer S, Stevens DA. Application of DNA typing meth- 29 Koç AN, Kocagoz S, Erdem F, Gundez Z. Outbreak of
ods to epidemiology and taxonomy of Candida Species. nosocomial fungemia caused by C.glabrata. Mycoses 2002;
J Clin Microbiol 1987; 25: 675 9. 45: 11 2.
15 Delabesse EB, Boiron P, Carlotti A, Dupont B. Candida 30 Cormican MG, Hollis RJ, Pfaller MA. DNA macrorestric-
albicans genotyping in studies with patients with AIDS tion profiles and antifungal susceptibility of C. glabrata.
developing resistance to Fluconazole. J Clin Microbiol Diagn Microbiol Infect Dis 1996; 25: 83 7.
1993; 31: 2933 7. 31 Barrchiesi F, Falconi Di Francesco L, Arzeni D, caselli F,
16 Clemons KV, Feroze F, Holmbreg K, Stevens DA. Com- Gallo D, Scalise G. Electrophoretic karyotyping and triaz-
parative analyses of genetic variability among Candida ole susceptibility of Candida glabrata clinical isolates. Eur
albicans isolates from different geographic locales by three J Clin Microbiol Infect Dis 1999; 18: 184 7.
genotypic methods. J Clin Microbiol 1997; 35: 1332 6. 32 Bennett JE, Izumikawa K, Marr KA. Mechanism of
17 Vasquez JA, Dembry LM, Sanchez V. Nosocomial Candida increased fluconazole resistance in Candida glabrata during
glabrata colonization: an Epifemiologic Study. J Clin prophylaxis. Antimicrob Agents Chemother 2004; 48:
Microbiol 1998; 36: 421 6. 1773 7.
18 Arif S, Barkham T, Power EG, Howell SA. Techniques for 33 Jong HS, Myung JC, Jeong WS, et al. Changes in karyotype
investigation of an apparent Outbreak of infections with and azole susceptibility of sequential bloodstream isolates
Candida glabrata. J Clin Microbiol 1996; 34: 2205 9. from patients with Candida glabrata candidemia. J Clin
19 Vasquez JA, Beckley A, Donabedian S, Sobel JD, Zervos MJ. MIcrobiol 2007; 45: 2385 91.
Comparaison of restriction enzyme analysis vs. pulsed field 34 Klempp-Selb B, Rimek D, Kappe R. Karyotyping of C.
gradient gel electrophoresis as a typing system for Torul- albicans and C. glabrata from patients with Candida sepsis.
opsis glabrata and Candida species other than C.albicans. Mycoses 2000; 43: 159 63.
J Clin Microbiol 1993; 31: 2021 30. 35 Lockhart SR, Fritch JJ, Sturdevant MA et al. Colonis-
20 Soll DR. The ins and outs of DNA fingerprinting. Clin ing populations of Candida albicans are clonal in origin
Microbiol Rev 2000; 13: 332 70. but undergo microevolution through C1 fragment
21 Fujita S, Hashimoto T. DNA fingerprinting patterns of reorganisation as demonstrated by DNA fingerprinting
Candida species using HinfI endonuclease. Int J Syst Evol and C1 sequencing. J Clin Microbiol 1995; 33:
Microbiol 2000; 50: 1381 9. 1501 9.
22 Williams JH, Kubelik AR, Rafalski JA, Tingey SV. DNA 36 Schmid J, Voss E, Soll DR. Computer-Assisted Methods for
polymorphisms amplified by arbitrary primers are assessing strain relatedness in Candida albicans by finger-
useful as genetis markers. Nucleic Acids Res 1990; printing with the moderately repetitive sequence Ca3.
18: 6531 5. J Clin Microbiol 1990; 28: 1236 43.
23 Robert F, Lebreton F, Bougnoux ME et al. Use of random 37 Bougnoux ME, Morand S, dÕenfert C. Usefulness of
amplified polymorphic DNA as a typing method for multilocus sequence typing for characterization of clinical
Candida albicans in epidemiological surveillance of burn isolates of Candida albicans. J Clin Microbiol 2002; 40:
unit. J Clin Microbiol 1995; 33: 2366 71. 1290 7.
24 Becker K, Badehorn D, Deiwick S, Peters G, Fegeler W. 38 Dodgson AR, Pujol C, Pfaller MA, Denning DW, Soll DR.
Molecular genotyping of Candida species with special Evidence for recombination in Candida glabrata. Fungal
respect to C. glabrata strains by arbitrarily primers PCR. Genet Biol 2005; 42: 233 43.
J Med Microbiol 2000; 49: 575 81. 39 Bougnoux M, Tavanti A, Bouchier C et al. Collaborative
25 Schwab U, Chernomas F, Larcon L, Weems J. Molecular consensus for optimised multilocus sequence typing
typing and fluconazole susceptibility of urinary Candida of Candida albicans. J Clin Microbiol 2003; 41: 5265 6.
glabrata isolates from hospitalized patients. Diagn Microbiol 40 Tavanti A, Neil ARG, Maiden MCJ, Odds FC, Shaw JD.
Infect Dis 1997; 29: 11 7. Genetic evidence for recombination in Candida albicans
26 Bougnoux ME, Robert F, Beria S, Cassinat B, Nicolas MH, based on haplotype analysis. Fungal genet biol 2004; 41:
Dupoy-Camet J. Typage moléculaire dÕisolats de Candida 553 62.
par «Random Amplified Polymorphic DNA. J Mycol Med 41 Hamel P, Kappe R, Rimek D. Rate of transmission and
1994; 4: 3 8. endogenous origin of Candida albicans and Candida glabrata
27 Chen KW, Lo HJ, Lin YH, Li SY. Comparison of four on adult intensive care units studied by pulsed field gel
molecular typing methods to assess genetic relatedness of electrophoresis. J Hosp Infect 2001; 49: 37 42.
Candida albicans clinical isolates in Taiwan. J Med Microbiol 42 Bougnoux ME, Cac G, Aegepter P, dÕenfert C, Fagon JY,
2005; 54: 249 58. candiRea study group. Candidemia and candiduria in
28 Di Francesco LF, Barchiesi F, Caselli F, Cirioni O, Scalise G. critically ill patients admitted to intensive care units
Comparison of four methods for DNA typing of clinical in France: incidence, molecular diversity, management
isolates of Candida glabrata. J Med Microbiol 1999; 48: and outcome. Intensive Care Med 2008; 34: 292 9.
955 63.
Ó 2009 Blackwell Verlag GmbH " Mycoses 53, 463 467 467