5830_04_p232-238

MICROBIAL DRUG RESISTANCE
Volume 11, Number 3, 2005
Mary Ann Liebert, Inc.
Establishing Surrogate Markers for Fluconazole ABSTRACT
Azole-resistant Candida can be a confounding factor for clinical management of opportunistic infections in
immunocompromised patients, but rapid identification of such resistant organisms can improve patient out-
come. New target-based molecular diagnostic strategies have the potential to identify resistant organisms faster
than current culture-based assays. It was the objective of this study to determine whether target site muta-
tions and/or drug pump over-expression are suitable surrogate markers of drug resistance that could aid new
molecular-based diagnostic assays. A collection of 59 clinical isolates displaying a range of azole susceptibili-
ties were assayed for mutations within the target gene Erg11 and for over-expression of drug-efflux pumps
Cdr1, Cdr2, Flu1, and Mdr1, as well as drug target gene Erg11 by quantitative real-time PCR with molecu-
lar beacons. A fluconazole-resistant (MIC Ն 64 ␮g/ml) phenotype was closely associated with over-expression
of Cdr1 (p ؍ 0.005), Cdr2 (p ؍ 0.01), and Mdr1 (p ؍ 0.03) along with four mutations in Erg11 (T229A, Y132F,
S405F, G464S). Changes in expression levels for Erg11 and Flu1 were not statistically correlated with resis-
tance (p ؍ 0.27 and p ؍ 0.86, respectively). Overall, these findings provide a statistical basis to establish Erg11
mutations and drug pump over-expression as surrogate markers for phenotypic fluconazole resistance.
INTRODUCTION
dients as driving forces and have more limited substrate speci-ficity. In C. albicans, the ABC transporter genes Cdr1 and Cdr2 TRIAZOLE ANTIFUNGAL AGENTS (e.g., fluconazole, itracona- and MFS gene Mdr1 have been linked to clinical drug resis-
zole, etc.) are commonly used to treat Candida infections tance.4–6,11,16–18,20–22 A novel MFS transporter gene Flu1 has because of their high therapeutic index. Resistance to these been identified in C. albicans, but its contribution to clinical agents, which were introduced in the early 1990s, is most no- table among immunosuppressed patients.5,13 The molecular In an effort to identify potential surrogate markers for clin- mechanisms of triazole resistance in C. albicans have been well ical drug resistance, the genes responsible for diverse resistance characterized in recent years and include alteration of the tria- mechanisms were evaluated in a collection of clinical isolates zole drug target enzyme, lanosterol 14␣-demethylase encoded displaying fluconazole susceptible (S), susceptible dose-depen- by Erg11, and reduction in cellular drug accumulation.5,6,17,21,22 dent (SDD), and resistant (R) in vitro susceptibility phenotypes Alterations in Erg11p include mutations that result in reduced by DNA sequencing and molecular beacon-based transcrip- affinity to triazole antifungal drugs, as well as those that lead to over-expression of Erg11, which diminishes drug efficacyvia target titration.4–6,8,15,16,22 A reduction of cellular drug ac-cumulation has also been linked to over-expression of multidrug MATERIALS AND METHODS
resistance (MDR) efflux transporter genes of the ATP-bindingcassette (ABC) and the major facilitator superfamily (MFS) Strains and susceptibility testing classes.4–6,11,14,16–18,20–22 ABC transporters require energy inthe form of ATP and can expel a multitude of substrates in con- Candida albicans strains used in this study include a col- trast to MFS transporters that use electrochemical proton gra- lection of 59 fluconazole-susceptible (MIC Յ 8.0 ␮g/ml) (n ϭ Public Health Research Institute, Newark, NJ 07103.
MARKERS FOR FLUCONAZOLE RESISTANCE IN C. albicans
14), susceptible dose-dependent (MIC 16–32 ␮g/ml) (n ϭ 13), coding region was performed using the DTCS Quick Start kit and resistant clinical isolates (MIC Ն 64 ␮g/ml) (n ϭ 32) as (Beckman Coulter, Fullerton, CA) and a CEQ 2000 XL capil- determined by the microdilution antifungal susceptibility test- lary electrophoresis DNA sequencer (Beckman Coulter).
ing method outlined by the National Committee for ClinicalLaboratory Standards (NCCLS) document M27-A.9 These Primer and molecular beacon design for strains were obtained from separate patients from diverse geo- graphic regions and were not epidemiologically linked. Flu- PCR primers and molecular beacon probes were designed conazole was obtained from Pfizer (New York).
to amplify and identify C. albicans genes Cdr1, Cdr2, Erg11,Flu1, Mdr1, and Pma1 as described by the molecular beacons synthesis protocol (http//:www.molecular-beacons.org). Each Candida albicans strains were grown in YPD medium molecular beacon was labeled with tetrachloro-6-carboxylflu- (Difco, Franklin Lakes, NJ) at 30°C until mid-log phase growth.
orescein (TET) (BioSearch Technologies, Inc., Novato, CA).
Total DNA-free RNA was isolated using the RNeasy Minikit Primer and molecular beacon probe sequences are listed in (Qiagen, Valencia, Calif.). For DNA extraction, an overnight Table 1. For each probe, a linear relationship was observed culture grown at 30°C was used and total genomic DNA was between the cycle threshold (Ct) value and the logarithm of extracted using the Wizard genomic DNA purification kit known amount of initial template molecules (not shown). Fur- thermore, each molecular beacon was shown to be specific toits target and did not show any cross-interaction or nonspe- PCR and sequencing primers were designed to the coding region of Erg11 (GenBank accession number X13296) usingOligo 4.04 primer design software (Molecular Biology Insights, First-strand cDNA was synthesized using Omniscript RT Inc., Cascade, CO). Primers used for PCR (5Ј-3Ј) were F52 PCR Kit (Qiagen) and 1 ␮g of total RNA. Samples incubated (GACAAAGAAAGGGAATTCAATCG) and R1763 (CACT- in the absence of reverse transcriptase were used as controls.
GAATCGAAAGAAAGTTGCCG). PCR amplification was Quantitative real-time PCR mixtures contained 0.2 ␮M of each carried out at 94°C for 1 min followed by 30 cycles of 94°C molecular beacon, 10 pmol of each primer, 2 U of AmpliTaq for 30 sec, 55°C for 30 sec, and 72°C for 2 min and a final ex- Gold DNA polymerase (Applied Biosystems, Foster City, CA), tension step of 72°C for 5 min. Primers used for DNA se- 0.25 mM dNTPs, 4 mM MgCl2, 1ϫ PCR buffer, and 250 ng quencing included F478 (GATGTTTCTGCTGAAGATGC), of cDNA. The thermal cycling program consisted of 10 min on F947 (GTGGTGATATTGATCCAAATCG), F1363 (CCAG- a spectrofluorometric thermal cycler (Applied Biosystems 7700 GTTATGCTCATACTAG), R498 (GCATCTTCAGCAGAA- Prism) at 95°C followed by 40 cycles of 20 sec at 95°C, 30 sec ACATC), R968 (GATTTGGATCAATATCACCAC), R1382 at 50°C, and 30 sec at 72°C, and real-time fluorescence was (CTAGTATGAGCATAACCTGG), and the PCR primers as monitored during the reaction and plotted against the number described above. Automated DNA sequencing of the Erg11 of thermal cycles. All quantitative real-time PCR assays were PRIMERS AND MOLECULAR BEACON DESIGN FOR QUANTITATIVE REAL-TIME PCR Sequence (5Ј–3Ј) aConstitutively expressed internal control gene.
SUMMARY OF CANDIDA albicans STRAINS USED IN THIS STUDY FLU resistance gene over-expressionc aOP, Oro-pharynx.
bS, Susceptible (MIC Յ 8.0 ␮g/ml); SDD, susceptible dose dependent (MIC 16–32 ␮g/ml); R, resistant (MIC Ն 64 ␮g/ml).
cOver-expression is defined as Ͼ 3 times the average FLU-susceptible strains values.
dRelative to GenBank Accession number X13296.
MARKERS FOR FLUCONAZOLE RESISTANCE IN C. albicans
performed in triplicate and the essential gene Pma1 was used ent in fluconazole-susceptible strains (Table 2). The mutation as a highly-expressed internal normalization marker.
S405F was only found in SDD and resistant strains, while othermutations also occurred in strains with SDD phenotypes (Table 2). Nine isolates had multiple amino acid mutations in Erg11.
Association of individual mdr gene expression levels for Cdr1, Cdr2, Mdr1, Erg11, and Flu1 relative to azole-suscepti- Quantitative over-expression of Erg11 and ble, susceptible dose-dependent and resistant phentoypes was determined by one factor analysis of variance (ANOVA). p val- Molecular beacon probes were used to quantify the level of ues less than 0.05 were considered significant.
expression of azole target gene Erg11, and mdr pumps Cdr1,Cdr2, Flu1, and Mdr1. The 14 azole-susceptible strains wereused to establish baseline values of transcripts for expressionof the known resistance genes Cdr1, Cdr2, Erg11, Flu1, and Mdr1. As expected, a majority of the susceptible strains dis-played low levels of expressed mdr genes, which remained closely grouped as shown in the scatter plot diagram (Fig. 1).
An epidemiologically diverse collection of 59 clinical iso- The susceptible dose-dependent isolates similarly displayed lates of C. albicans encompassing fluconazole susceptible, low-level expression of mdr genes, with the exception of Mdr1, dose-dependent, and resistant phenotypes were evaluated by which revealed moderate to high increases (2- to 10-fold) of DNA sequence analysis of Erg11. Numerous mutations rela- transcript. The azole-resistant isolates showed a 2- to 45-fold tive to the published Erg11 sequence (GenBank accession num- increase in Mdr1 transcript levels when compared to the sus- ber X13296) were identified resulting in 15 different amino acid ceptible and susceptible dose-dependent isolates. For Cdr1, all substitutions, as follows: D116E, F126L, K128T, Y132F, of the susceptible dose-dependent isolates had similar expres- K143R, T229A, E266D, S405F, V437I, D446N, F449S, sion profiles as the susceptible. However, Cdr1 over-expres- G450E, G464S, R467K, and V488I (Table 2). The mutations sion was found in ϳ34% (11/32) of the total R isolates tested ranged in frequency from 2.2 to 8.8% in the 45 R or SDD strains with 2- to 15-fold increases in transcript levels. A majority of (Table 3) and clustered within four mutational “hot-spot” re- the isolates also co-expressed Cdr2 (n ϭ 11), supporting the gions previously identified, which were located between amino previous notion that Cdr1 and Cdr2 are co-regulated.21 Over- acid residues 72 to 467.8,12 However, six amino acid mutations, all, the frequency of gene over-expression for Cdr1, Cdr2, and D116E, K128T, G450E, V437I, D446N, and F449S, were not Mdr1 was 24.4%, 24.4%, and 33.3%, respectively, for all R and linked to triazole resistance as they were also found to be pres- SDD isolates (n ϭ 45) examined (Table 3). Gene over-expres- PREVALENCE AND ASSOCIATION OF SPECIFIC GENETIC MECHANISMS WITH RESISTANCE aANOVA analyses were performed on the absolute expression levels values of the mdr genes.
bOver-expression is defined as Ͼ3 times the average FLU susceptible strains values.
cND, Not determined because population size was too limited for statistical significance.
PARK AND PERLIN
Relative expression levels of mdr genes. Quantitative real-time PCR with molecular beacons of known mdr genes was performed on cDNA transcripts of a panel of C. albicans isolates (n ϭ 59). The collection included 14 azole-susceptible (ٗ), 13azole-susceptible dose-dependent (X), and 32 resistant (᭝) isolates. Variations in the initial amount of mRNA among the dif-ferent samples were normalized to the level of Pma1 mRNA, yielding a ratio between the specific mdr gene cDNA copy num-ber and Pma1 cDNA copy number.
sion was constitutively expressed in the strains, because there level resistance largely results from incremental susceptibility was no evidence for significant changes in expression due to shifts mediated by resistance mechanisms induced in response drug exposure. Erg11 over-expression was found in ϳ12% of to short- or long-term drug exposure.4,7,21 Nevertheless, there the resistant isolates tested (4/32) whereas Flu1 over-expres- are subsets of the various resistance mechanisms that are more sion was found in one isolate, but neither Erg11 nor Flu1 over- frequently associated with in vitro resistance and therefore have expression were statistically linked to clinical drug resistance the potential to serve as diagnostic surrogates. Advances in mol- (p ϭ 0.27 and p ϭ 0.87, for Erg11 and Flu1, respectively) ecular diagnostic approaches have made it possible to rapidly evaluate genetic changes (allele discrimination and gene over- Five azole-susceptible dose-dependent strains had single expression) in an organism that result in a prominent pheno- mechanisms of resistance in contrast to 20 triazole-resistant iso- type, such as drug resistance.1,3,10 Most recently, it was shown lates that displayed multiple mechanisms of resistance (ϳ62%).
that gene over-expression of multiple resistance targets could A majority of strains showed multiple types (target site or drug be profiled in a simple real-time assay,3 and at least seven dif- efflux pump) of resistant mechanisms (Table 2). Of 10 isolates ferent alleles associated with itraconazole resistance in As- with a single resistance mechanism, five had target-based mu- pergillus fumigatus could be readily detected in a single mul- tations and five showed efflux pump over-expression. Of the strains with target site mutations linked to resistance, a major- In this report, the most commonly associated resistance ity also showed over-expression of mdr pumps.
mechanisms include Erg11 target site mutations T229A,Y132F, K143R, E266D, S405F, G464S, and V488I (Table 2)and over-expression of drug efflux transporter genes Cdr1, DISCUSSION
Cdr2, and Mdr1, above a baseline threshold value of 3-foldgreater than the average expression in susceptible strains (Fig.
It is now well established that traizole resistance in C. albi- 1). In each case, the manifestation of either a given Erg11 mu- cans is multifactorial involving several different resistance tation and/or gene over-expression was correlated with resis- mechanisms operating individually or collectively in a given tance. All the Erg11 mutations have been identified previ- strain.4,7,11,16,21,22 Such complexity implies that no one resis- ously.4,6–8,15,16 Y132F is one of the most frequent Erg11 tance mechanism plays a dominant role. Furthermore, high- mutations encountered (Table 3). This residue is believed to re- MARKERS FOR FLUCONAZOLE RESISTANCE IN C. albicans
side close to the access channel for azoles in the demethylase Overall, Ͼ90% of the 32 resistant strains could be accounted enzyme, which may alter entry of the inhibitor to its binding for by target site mutations and/or up-regulation of drug efflux site.8,12 Mutation T229A was found unlinked to other resistance transporters. Only three strains, 31, 32, and 57 (containing sta- mechanisms. It is believed to reside on the F helix, which may tistically unlinked D116E, E266D mutations) had unknown contribute dynamically to the access channel for drug. Al- mechanisms. These results suggest that existing mechanisms though, independent confirmation of the contribution of this provide excellent coverage for non-culture-based assays of flu- mutation to resistance has not been reported. Amino acid sub- conazole resistance. As new targets emerge and are validated, stitutions involving E266D and V488I mutation are not likely they can be incorporated into the new generation of nucleic acid to be important for high-level resistance because they have been amplification systems with self-reporting probes capable of al- observed in susceptible strains.8 Residue G464, which is highly lele discrimination and quantitative assessment of target num- conserved across all known species, appears to lie directly be- hind the plane of the heme prosthetic. The G464S mutation mayinterfere with coordination of the azole molecule. Finally, theS405F substitution lies immediately beyond the K helix near ACKNOWLEDGMENTS
the substrate- and azole-binding pocket.8 The dynamic natureof the demethylase enzymes helps account for the numerous The authors acknowledge the following groups for provid- mutations that can confer varying levels of resistance and may ing clinical isolates: Phyllis Della-Latta (Columbia-Presbyter- result in either resistant or intermediate-level resistance such as ian Medical Center, New York), Brahm H. Segal (Roswell Park S405F or can result in common mutations such as D116E, Cancer Institute, New York), Theodore C. White (Seattle Bio- which have a weak or no apparent phenotype. Mutations F126L medical Research Institute). Joachim Morschhäuser (Univer- and K143R are often found together along with up-regulation sität Würzburg, Germany), Frank Odds (University of Ab- of an efflux transporter, suggesting that their effect is weak.
erdeen, UK), Spencer Redding, Jose Lopez-Ribot, and Thomas Overall, it appears that mutations T229A, Y132F, S405F, and F. Patterson (University of Texas Health Science Center at San G464S are sufficient to confer fluconazole resistance. Their fre- Antonio, Texas). The authors are also grateful to Salvatore quency in the clinical resistant population is difficult to assess A. E. Marras and Barun Mathema for helpful discussions of because the sample size in this study was too small to deduce molecular beacon design and statistical analyses.
statistic significance. Nevertheless, these alleles are likely to bereliable indicators that could serve as useful clinical markers inscreening for triazole resistance.
The over-expression of multidrug efflux transporters repre- REFERENCES
sents the second major mechanism associated with resis-tance.11,16,17,20–22 The C. albicans genome contains in excess 1. Balashov S.V., R. Gardiner, S. Park, and D.S. Perlin. 2005.
of 100 ORFs encoding putative ABC- and mfs-type trans- Rapid, high-throughput, multiplex, real-time PCR for identification porters. In principle, if expressed at a suitable level, each gene of mutations in the cyp51A gene of Aspergillus fumigatus that con-
fer resistance to itraconazole. J. Clin. Microbiol. 43:214–222.
product has the potential to export azole antifungals actively 2. Calabrese, D., J. Bille, and D. Sanglard. 2000. A novel multidrug
out of the cell. Yet, most studies of clinical isolates with resis- efflux transporter gene of the major facilitator superfamily from tant properties appear to be limited to the over-expression of Candida albicans (FLU1) conferring resistance to fluconazole Mi- relatively few transporter genes including Mdr1, Cdr1, and crobiology. 146:2743–2754.
Cdr2, which were identified in this study as being closely linked 3. Frade, J.P., D.W. Warnock, and B.A. Arthington-Skaggs. 2004
to resistance (p Ͻ 0.05; Table 3). The up-regulation of mdr Rapid quantification of drug resistance gene expression in Candida genes is a common cellular response to different types of stress.
albicans by reverse transcriptase LightCycler PCR and fluorescent For this reason, establishing a baseline with susceptible isolates probe hybridization. J. Clin. Microbiol. 42:2085–2093.
was critical to assign threshold values for specific gene over- 4. Franz, R., S.L. Kelly, D.C. Lamb, D.E. Kelly, M. Ruhnke, and
expression, which allows a more reliable association with re- J. Morschhauser. 1998. Multiple molecular mechanisms con-
tribute to a stepwise development of fluconazole resistance in clin-
sistance (Fig. 1). It is also the first step toward using such mea- ical Candida albicans strains Antimicrob Agents Chemother.
surements as a diagnostic indicator. Only one isolate showed 42:3065–3072.
any over-expression of Flu1, suggesting that this transporter 5. Franz, R., M. Ruhnke, and J. Morschhauser. 1999. Molecular
does not contribute appreciably to clinical resistance. Further- aspects of fluconazole resistance development in Candida albicans more, Erg11 over-expression also showed a poor correlation Mycoses. 42:453–458.
(p ϭ 0.27) with resistance despite a wide range in expression 6. Loeffler, J., and D.A. Stevens. 2003. Antifungal drug resistance.
levels. The molecular mechanisms responsible for azole resis- Clin Infect Dis 36:S31–S41.
tance in C. albicans are common to other Candida spp., as well 7. Lopez-Ribot, J. L., R.K. McAtee, S. Perea, W.R. Kirkpatrick,
as to less related fungi.4,5,8,21 The evolution of drug resistance M.G. Rinaldi, and T.F. Patterson. 1999 Multiple resistant phe-
in a susceptible strain may involve a single mechanism or may notypes of Candida albicans coexist during episodes of oropha-ryngeal candidiasis in human immunodeficiency virus-infected pa- reflect a step-wise accumulation of mutations resulting in tar- tients. Antimicrob. Agents Chemother. 43:1621–1630.
get site affinity changes8,12 and/or induction of various types 8. Marichal, P., L. Koymans, S. Willemsens, D. Bellens, P. Ver-
of drug efflux transporters.4,7,20,22 The combination of resis- hasselt, W. Luten, M. Borgers, F.C.S. Ramaekers, F.C. Odds,
tance mechanisms seems to be associated with a high level of and H. Vanden Bossche. 1999. Contribution of mutations in the
azole resistance in C. albicans, and an effective diagnostic must cytochrome P450 14a-demethylase (Erg11p, Cyp51p) to azole re- be able to scan multiple types of mechanisms.
sistance in Candida albicans. Microbiology 145:2701–2713.
PARK AND PERLIN
9. National Committee for Clinical Laboratory Standards. 1997.
16. Sanglard, D., F. Ischer, D. Calabrese, M.de Micheli, and J. Bille.
Reference method for broth dilution antifungal susceptibility test- 1998. Multiple resistance mechanisms to azole antifungals in yeast ing of yeasts. Approved standard, NCCLS document M27-A. Na- clinical isolates. Drug Res. Updates 1:255–265.
tional Committee for Clinical Laboratory Standards, Wayne, PA.
17. Sanglard, D., and F.C. Odds. 2002. Resistance of Candida species
10. Perlin, D.S., and S. Park. 2001. Rapid identification of fungal
to antifungal agents: molecular mechanisms and clinical conse- pathogens: molecular approaches for a new millennium. Rev. Med.
quences. Lancet Infect Dis 2:73–85.
Microbiol. 12:S13–S20.
18. Sheehan, D. J., C.A. Hitchcock, and C.M. Sibley. 1999. Current
11. Perea, S., J.L. Lopez-Ribot, W.R. Kirkpatrick, R.K. McAtee,
and emerging azole antifungal agents. Clin Microbiol Rev.
R.A. Santillan, M. Martinez, D. Calabrese, D. Sanglard, and
12:40–79.
T.F. Patterson. 2001. Prevalence of molecular mechanisms of re-
19. Tyagi, S., D.P. Bratu, and F.R. Kramer. 1998. Multicolor molec-
sistance to azole antifungal agents in Candida albicans strains dis- ular beacons for allele discrimination. Nature Biotechnol. 16:49–53.
playing high-level fluconazole resistance isolated from human 20. White, T.C. 1997. Increased mRNA levels of ERG16, CDR, and
immunodeficiency virus-infected patients Antimicrob. Agents MDR1 correlate with increases in azole resistance in Candida al- Chemother. 45:2676–2684.
bicans isolates from a patient infected with human immunodefi- 12. Podust, L. M., T.L. Poulos, and M.R. Waterman. 2001. Crystal
ciency virus. Antimicrob. Agents Chemother. 41:1482–1487.
structure of cytochrome P450 14-sterol demethylase (CYP51) from 21. White, T.C., K.A. Marr, and R.A. Bowden. 1998 Clinical, cel-
Mycobacterium tuberculosis in complex with azole inhibitors. Proc.
lular, and molecular factors that contribute to antifungal drug re- Natl. Acad. Sci. USA 98:3068–3073.
sistance. Clin. Microbiol. Rev. 11:382–402.
13. Safdar, A., V. Chaturvedi, E.W. Cross, S. Park, E.M. Bernard,
22. White, T.C., S. Holleman, F. Dy, L.F. Mirels, and D.A. Stevens.
D. Armstrong, and D.S. Perlin. 2001. Prospective study of Can-
2002. Resistance mechanisms in clinical isolates of Candida albi- dida species in patients at a comprehensive cancer center Antimi- cans. Antimicrob. Agents Chemother. 46:1704–1713.
crob. Agents Chemother. 45:2129–2133.
14. Sanglard, D., F. Ischer, M. Monod, and J. Bille. 1997. Cloning
of Candida albicans genes conferring resistance to azole antifun- gal agents: characterization of CDR2, a new multidrug ABC trans- porter gene. Microbiology 143:405–416.
at the International Center for Public Health 15. Sanglard, D., F. Ischer, L. Koymans, and J. Bille. 1998. Amino
acid substitutions in the cytochrome P-450 lanosterol 14alpha- demethylase (CYP51A1) from azole-resistant Candida albicans clinical isolates contribute to resistance to azole antifungal agents.
Antimicrob Agents Chemother 42:241–253.

Source: http://www.molecular-beacons.org/download/park,mdr05(11)232.pdf