RT-PCR vs microarrays for measuring expression of antibiotic resistance genes

Please cite this article as follows: PCR Encyclopedia (2005) 1: 092249-07

RT-PCR vs microarrays for measuring expression of antibiotic resistance genes

PCR Encyclopedia (2005) 1: 092249-07

Utilization of DNA microarrays and RT-PCR for measuring the expression of antibiotic resistance genes. What should be better suited for its implementation in clinical laboratories?

Juan F. Linares¹, Victor Parro², José L. Martínez¹,*, and Ingegerd Gustafsson^1,3

1 Departamento de Biotecnología Microbiana. Centro Nacional de Biotecnología (CSIC), Darwin 3, Campus UAM, Cantoblanco, 28049-Madrid, Spain
2Laboratorio de Ecologia Molecular, Centro de Astrobiologia (CSIC-INTA), Carretera de Ajalvir km 4, 28850 Torrejon de Ardoz, Madrid, Spain
3Present address: Clinical Microbiology, County hospital, SE-301 85 Halmstad, Sweden.

*Corresponding author

Abstract
Background: Bacterial antibiotic resistance can be achieved by acquisition of antibiotic resistance genes, by mutation in target genes, or by overproduction of weakly-expressed antibiotic resistance genes (c. a. chromosomal beta-lactamases or multidrug efflux pumps) present in the chromosomes of all isolates of a given bacterial species. Whereas the molecular methods for tracking the presence antibiotic resistance genes or target mutations are well established, there is a general lack of reliable methods to be used in clinical laboratories for the analysis of the expression of antibiotic resistance genes.
Methods: Mutants overexpressing the best characterized multidrug efflux pumps of Pseudomonas aeruginosa were obtained by single-step selection on plates containing selective antibiotics. A DNA microarray was constructed containing PCR fragments of 12 genes in total; the cytoplasmic pumps (MexB, MexD, MexF, MexY), the fusion proteins (MexA, MexC, MexE, MexX), the regulatory proteins (MexR, NfxB, MexT, MexZ) and one control DNA. The probes attached to the glass slide consisted of 300 to 1000-bp PCR products. Differential expression of the MDR pumps was estimated in the mutants and wild-type strains by semiquantitative RT-PCR and cDNA hybridization to the DNA-microarray
Results: Two feasible methods for analyzing overexpression of P. aeruginosa MDR pumps in clinical microbiology laboratories were established. Both Semiquantitative RT-PCR and a PCR fragments-based DNA microarray reliabily distinguished MDR overproducers from the wild-type strain. Whereas RT-PCR is easier to implement at clinical laboratories, the DNA-microarray technique allows the analysis of a larger number of antibiotic resistance genes in a single assay.
Conclusions: New feasible and reliable techniques are required for analyzing expression of antibiotic resistance genes in clinical microbiology laboratories. Using well-defined P. aeruginosa antibiotic resistant mutants expressing specific MDR efflux pumps, we have tested two possible methods. Both were feasible and reliable. Whereas, semiquantitative RT-PCR was easier to implement, DNA-microarray hybridization allowed the analysis in a single experiment of a larger number of genes.

Introduction
Molecular epidemiology of antibiotic resistance mainly relies in methods such as Southern blotting and the polymerase chain reaction (PCR) designed to detect the presence of antibiotic resistance genes [1]. More recently, DNA-microarrays have proven to be a good alternative for the simultaneous search of different antibiotic resistance determinants in a single experiment [2, 3]. Beside that, the presence of specific mutations in target genes has been analyzed using combinations of PCR and sequencing [4, 5], and the utilization of DNA-microarrays for detecting single nucleotide polymorphisms (SNPs) might also be a potential fast alternative for these studies.
Those methods are useful either for detecting the presence of antibiotic resistance genes or for detecting mutations in target genes. Nevertheless, they are useless in the case of genes, like multidrug (MDR) efflux pumps, that are present in all isolates of a given bacterial species and contribute to antibiotic resistance when they are overexpressed [1, 6]. The molecular epidemiology of this type of antibiotic resistance must then rely on methods for measuring the level of expression and not the presence or absence of antibiotic resistance genes. Northern blotting may be a good alternative for these analyses. However, RNAs from bacteria have short lifetimes and this technique is not easy to implement in clinical settings. Another technique that might be useful could be RT-PCR. The utilization of DNA-microarrays might also be a good approach when the expression of several genes is tested at the same time in a single experiment. If fact, DNA-microarrays have been mainly built for transcriptomic analysis and the study of transcription of antibiotic resistance genes would just be a specific application of this wider type of analysis.
As a model for studying the reliability of RT-PCR and DNA-microarray techniques for measuring the expression of antibiotic resistance genes, we have chosen Pseudomonas aeruginosa. This bacterial species is intrinsically resistant to a wide range of antibiotics due to the presence of several MDR pumps in its genome (http://www.pseudomonas.com). The four best-characterized MDR pumps of P. aeruginosa are MexAB-OprM [7], MexCD-OprJ [8], MexEF-OprN [9] and MexXY [10]. Expression of those MDR pumps are usually down-regulated by specific transcriptional regulators [11]. However, mutants that overexpress these pumps and have increased levels of resistance to several different antibiotics are easily obtained in vitro [12] and from infected patients under antibiotic treatment [13, 14]. In the present article we have designed and validated methods for measuring the expression of these pumps using as controls single-step spontaneous MDR mutants [15] that overproduce each of these four different efflux pumps.

Methods
Strains
P. aeruginosa PAO1 wild-type strain and single-step MDR-overproducing mutants were form our laboratory collection [15]. The strain JFL30 was selected by seeding PAO1 on LB plates containing 20 µg/ml of tetracycline. The mutant has a lower susceptibility to quinolones, aztreonam and cefpirome, but not to imipenem, a phenotype that is compatible with overexpression of MexAB-OprM. The mutant JFL28 was selected by platting PAO1 on LB plates containing 8 µg/ml of norfloxacin and 500 µg/ml of erythromycin and has a decreased susceptibility to cefpirome, and quinolones, but not to aztreonam. This phenotype is compatible with MexCD-OprJ overexpression. The mutant JFL05 was selected with 600 µg/ml chloramphenicol and was susceptible to imipenem, chloramphenicol and quinolones, a phenotype compatible with MexEF-OprN overexpression. Finally, strain JFL10 was selected on Mueller Hinton (MH) agar plates containing 4 µg/ml of gentamicin and 1 µg/ml of ofloxacin. The mutant was less susceptible to quinolones and gentamicin, a phenotype compatible with MexXY-OprM overproduction. Bacteria were routinely grown in Luria-Bertani broth at 37ºC.
Construction of a DNA-microarray for measuring expression of MDR pumps in P. aeruginosa
Chromosomal DNA of P. aeruginosa PAO1 was obtained with the Genome DNA Kit (Bio 101) according to the instructions of the manufacturer. DNA polymerase gel form (Biotools) was used for the amplification of the different genes to be included in the DNA-microarray (0.5 µM each specific primer (Table 1) and 100 ng of chromosomal DNA). The PCR reaction mixtures were incubated for 5 min at 95ºC and 30 cycles of 30 s at 95ºC, 30 s at 53ºC (for mexA, mexC, mexD, mexF, mexT and nfxB), 58ºC (for mexE, mexX, mexY, mexR and mexZ,), and 61ºC (for mexB), and 1 min at 72ºC, before finishing with a final 10 min elongation at 72ºC. 5% DMSO was added at the PCR reaction in the case of mexT, and nfxB.

The PCR products obtained were purified with QIAEX II Gel Extraction Kit (Qiagen), sequenced to assure their identity, cloned in the pGEMT-Easy vector (Promega), and PCR-amplified again using the primers T7 (5’-TAATACGACTCACTATAGGG-3’) and Sp6 (5’-GATTTAGGTGACACTATAG-3’). For preparation of the slides the PCR fragments were dried under vacuum and resuspended in spotting solution (TeleChem Int.) at a final concentration of 400 ng/µl. Spotting (2-2.5 nl/spot) was carried out on epoxy-modified slides (TeleChem Int.) with a MicrGridII arrayer (Genomic Solutions, BioRobotics) at 45 to 50% humidity. P. aeruginosa chromosomal DNA was included as a control for the normalization and eight replicas of each spot were included in the microarray.
Preparation of RNA and RT-PCR
Total RNA was obtained as described from liquid bacterial cultures using Tri Reagent LS (Molecular Research Center, Inc.). Residual DNA was removed by treatment with DNA-Free (Ambion). RT-PCR assays were performed using Ready-To-Go RT-PCR beads (Amersham), specific primers (50 pmol) for the each analyzed gene and two serial 10-fold dilutions (1 and 0.1 µg) of bacterial RNA to ensure a linear response. The reaction mixtures were incubated for 30 min at 42ºC, followed by 10 min at 95ºC and 30 cycles of 30 s at 95ºC, 30 s at 55ºC (for mexC and mexF), 57ºC (for mexA), 60ºC (for mexX and rpsL), and 1 min at 72ºC, and with a final 7 min elongation at 72ºC. To ascertain that no residual DNA was present in the RNA preparations, a PCR reaction was performed using the same primers and overall conditions, except that no reverse transcriptase was added. Expression of the rpsL gene served as an internal control that ensured that equal amounts of RNA were used in all of the RT-PCRs done.
cDNA synthesis and labelling
cDNA synthesis and labelling was performed from 20 µg of each RNA sample by incorporation aminoallyle-dUTP (Sigma) using the SuperScript™ First-Strand Synthesis System (Invitrogen, Life technologies) as indicated by the manufacturer. The cDNA products were purified with QIAquick PCR purification kit (QIAGEN) according to the manufacturer’s instructions and eluted with 50µl of phosphate buffer 4mM, pH 8.5. The cDNA concentration was measured in a NanoDrop-1000 spectrophotometer (NanoDrop Technologies). Afterwards, the cDNA was dried and the pellet was suspended in 9 µl of NaHCO3 0.1M, pH 9. The cDNA was labelled with either Cy3 or Cy5 (Mono-reaction dye pack, Amersham Bioscience) following manufacturer’s instructions. The samples were purified again and the incorporation of the fluorophores was measured in a NanoDrop-1000 spectrophotometer (NanoDrop Technologies).
Hybridization conditions
Before hybridisation the microarray slides were washed 2 min with 2xSSC containing 0.1% sarcosyl, and 2 min with 2xSSC. The DNA on the slide was denatured by incubating the array in distilled water at 95ºC for 2 min. Afterwards, the array was washed 2 min in ice-cold 100% ethanol and spun-dried in a centrifuge. Paired Cy3· and Cy5 labelled cDNAs from isogenic wild-type and MDR overproducing strains were mixed, dried and the pellet suspended in 10 µl hybridisation buffer (5% dextransulphate, 3xSSC, 1% SDS, 50% formamide and 5xDenhart’s solution). The samples were heat-denatured for 2 min in 96ºC, applied on a cover slip (Hybri-slips, Sigma) and the coverslip applied on the microarray slide. Hydridisation was carried on at either 50ºC or 65ºC overnight. After hybridisation the slides were washed in 2x SSC with 0.1% sarcosyl for 1 min, 3 min in 2x SSC and 5 min in 0.2x SSC and spun-. The microarrays were scanned with Scan Array Lite (Perkin Elmer) and analysed using Scan Array Express (Perkin Elmer).

Results and Discussion
Analysis of the expression of MDR pumps by semiquantitative RT-PCR
The level of expression of mexAB-oprM, mexCD-oprJ, mexEF-oprN and mexXY-oprM in the MDR overproducing mutants in comparison with the wild-type strains was analysed by semiquantitative RT-PCR [16] as described in Methods. To make that, RNAs obtained at the exponential growth phase, were used. As shown in Figure 1, the method was reliable and we were able to detect overproduction of each of the MDR pumps. The mexAB-OprM and mexXY pumps were expressed, although at low level in the wild-type strains, whereas we were unable to detect expression of neither mexCD-OprJ nor mexEF-OprN unless they were overproduced. These data fit with the fact that MexAB-OprM [17] and MexXY [18] contribute to intrinsic antibiotic resistance in P. aeruginosa, whereas such a role has not been demonstrated neither for MexCD-OprJ nor MexEF-OprN.

Figure 1 - Semiquantitave RT-PCR of MDR efflux pumps in wild-type and multidrug resistant P. aeruginosa isolates.
The amounts of mRNAs of the first genes of the four analyzed MDR pumps were estimated in wild-type and MDR mutant strains as described in Materials and Methods. Overexpressed genes are highlighted with squares.

Analysis of the expression of MDR pumps using a DNA-microarray
A DNA microarray containing the structural and regulator genes of the most studied MDR pumps form P. aeruginosa, MexAB-OprM, MexCD-OprJ, MexEF-OprN and MexXY-OprM was constructed. Eight dots of each gene were included together with eight more dots of chromosomal DNA for data normalization. To analyze the reliability of this method, we have used single-step MDR-overproducing mutants obtained in our laboratory and classified as overproducers of each of the aforementioned MDR pumps (see above). To detect overexpression of the pumps, RNA was obtained from each of the different mutants and its isogenic wild-type strain, converted to cDNA and differentially labelled (see methods). Mixtures of labelled cDNAs from the MDR overproducing mutant and its isogenic wild-type strain were hybridized with the DNA-microarray. RNAs were obtained from LB-growing bacteria at two growing phases: exponential and stationary. We were unable to obtain reliable data with RNAs obtained from stationary growth phase, probably because bacterial RNAs at this stage of growth are less stable than at exponential growth phase. However, we were able to detect MDR overexpression using exponential phase P. aeruginosa cultures (see below). No reliable date were obtained either when the hybridization temperature was 55ºC. Call et al. [19] developed a DNA microarray for identifying multiple tetracycline resistance genes in several Gram-negative species. They experienced an increased sensitivity with probes of 550-bp since a 25-mer probe array was insensitive to low-copy-number genes. Nevertheless, the utilization of PCR fragments for constructing the microarray increases sensitivity but reduces specificity. For this reason, the hybridization temperature was increased to 65ºC. In these conditions, we were able to detect overproduction of MDR pumps in each of the mutants (Table 2). The antibiotic susceptibility profile of each mutant fits with the overexpressed pump as detected using this DNA-microarray technique. Besides that, we observed that the expression genes coding the local transcriptional regulators of the MDR pumps were also increased for each specific MDR pump. It has been shown that mexR [20] autoregulates their expression. Our results fit with these data and suggest that expression of nfxB, mexT and mexZ is autoregulated as well.

Conclusions
Novel techniques are required for the molecular epidemiology of antibiotic resistance due to overexpression of antibiotic resistance genes ubiquitously present in bacterial chromosomes. In this line, we have developed a DNA-microarray for detecting hyperexpression of MDR efflux pumps in P. aeruginosa and have compared this methodology with semiquantitative RT-PCR. Hyperexpression of MDR pumps can be detected using each of the techniques. Semiquantitative RT-PCR is easy to be implemented in clinical laboratories, and would be the current method of choice. However, the expression of each independent MDR pump must be independently measured. On the other hand DNA-microarray technology is not easily accessible to most clinical microbiology laboratories. However, one single hybridization can detect overproduction of several different house-keeping antibiotic resistance determinants, which indicates that the method may be extremely useful in the future upon implementation of this technology in clinical laboratories. In the present article we have established the conditions for applying both methods for the detection of P. aeruginosa mutants that overexpress MDR pumps.

Acknowledgements
JFL was the recipient of a fellowship from MEC. IG was the recipient of a fellowship from The Swedish Society of Medicine. This work has been supported by grants GEN2001-4698-C05, GR/SAL/0795/2004, and BIO2004-00702.

References
1. A Sundsfjord, GS Simonsen, BC Haldorsen, H Haaheim, SO Hjelmevoll, P Littauer, KH Dahl: Genetic methods for detection of antimicrobial resistance. Apmis 2004, 112:815-37.
2. V Perreten, L Vorlet-Fawer, P Slickers, R Ehricht, P Kuhnert, J Frey: Microarray-based detection of 90 antibiotic resistance genes of gram-positive bacteria. J Clin Microbiol 2005, 43:2291-302.
3. AH van Hoek, IM Scholtens, A Cloeckaert, HJ Aarts: Detection of antibiotic resistance genes in different Salmonella serovars by oligonucleotide microarray analysis. J Microbiol Methods 2005, 62:13-23.
4. RA Walker, N Saunders, AJ Lawson, EA Lindsay, M Dassama, LR Ward, MJ Woodward, RH Davies, E Liebana, EJ Threlfall: Use of a LightCycler gyrA mutation assay for rapid identification of mutations conferring decreased susceptibility to ciprofloxacin in multiresistant Salmonella enterica serotype Typhimurium DT104 isolates. J Clin Microbiol 2001, 39:1443-8.
5. SS Richter, KP Heilmann, SE Beekmann, NJ Miller, CL Rice, GV Doern: The molecular epidemiology of Streptococcus pneumoniae with quinolone resistance mutations. Clin Infect Dis 2005, 40:225-35.
6. JL Martinez, F Baquero: Epidemiology of antibiotic-inactivating enzymes and DNA probes: the problem of quantity. J Antimicrob Chemother 1990, 26:301-3.
7. XZ Li, H Nikaido, K Poole: Role of mexA-mexB-oprM in antibiotic efflux in Pseudomonas aeruginosa. Antimicrob Agents Chemother 1995, 39:1948-53.
8. K Poole, N Gotoh, H Tsujimoto, Q Zhao, A Wada, T Yamasaki, S Neshat, J Yamagishi, XZ Li, T Nishino: Overexpression of the mexC-mexD-oprJ efflux operon in nfxB-type multidrug-resistant strains of Pseudomonas aeruginosa. Mol Microbiol 1996, 21:713-24.
9. T Kohler, M Michea-Hamzehpour, U Henze, N Gotoh, LK Curty, JC Pechere: Characterization of MexE-MexF-OprN, a positively regulated multidrug efflux system of Pseudomonas aeruginosa. Mol Microbiol 1997, 23:345-54.
10. T Mine, Y Morita, A Kataoka, T Mizushima, T Tsuchiya: Expression in Escherichia coli of a new multidrug efflux pump, MexXY, from Pseudomonas aeruginosa. Antimicrob Agents Chemother 1999, 43:415-7.
11. S Grkovic, MH Brown, RA Skurray: Regulation of bacterial drug export systems. Microbiol Mol Biol Rev 2002, 66:671-701.
12. A Alonso, E Campanario, JL Martinez: Emergence of multidrug-resistant mutants is increased under antibiotic selective pressure in Pseudomonas aeruginosa. Microbiology 1999, 145 ( Pt 10):2857-62.
13. L Pumbwe, LJ Piddock: Two efflux systems expressed simultaneously in multidrug-resistant Pseudomonas aeruginosa. Antimicrob Agents Chemother 2000, 44:2861-4.
14. I Ziha-Zarifi, C Llanes, T Kohler, JC Pechere, P Plesiat: In vivo emergence of multidrug-resistant mutants of Pseudomonas aeruginosa overexpressing the active efflux system MexA-MexB-OprM. Antimicrob Agents Chemother 1999, 43:287-91.
15. JF Linares, JA Lopez, E Camafeita, JP Albar, F Rojo, JL Martinez: Overexpression of the multidrug efflux pumps MexCD-OprJ and MexEF-OprN is associated with a reduction of type III secretion in Pseudomonas aeruginosa. J Bacteriol 2005, 187:1384-91.
16. A Alonso, JL Martinez: Expression of multidrug efflux pump SmeDEF by clinical isolates of Stenotrophomonas maltophilia. Antimicrob Agents Chemother 2001, 45:1879-81.
17. XZ Li, K Poole, H Nikaido: Contributions of MexAB-OprM and an EmrE homolog to intrinsic resistance of Pseudomonas aeruginosa to aminoglycosides and dyes. Antimicrob Agents Chemother 2003, 47:27-33.
18. N Masuda, E Sakagawa, S Ohya, N Gotoh, H Tsujimoto, T Nishino: Contribution of the MexX-MexY-oprM efflux system to intrinsic resistance in Pseudomonas aeruginosa. Antimicrob Agents Chemother 2000, 44:2242-6.
19. DR Call, MK Bakko, MJ Krug, MC Roberts: Identifying antimicrobial resistance genes with DNA microarrays. Antimicrob Agents Chemother 2003, 47:3290-5.
20. P Sanchez, F Rojo, JL Martinez: Transcriptional regulation of mexR, the repressor of Pseudomonas aeruginosa mexAB-oprM multidrug efflux pump. FEMS Microbiol Lett 2002, 207:63-8.

The PCR Encyclopedia