Optimization and Evaluation of Triplex Real-time PCR Assay for Detection of Genes Encoding Staphylococcal Virulence and Methicillin Resistance Using Two Different Multi-channel Emission Instruments

Aim of Study: To optimize a triplex real-time PCR assay developed elsewhere and to evaluate the performance characteristics of two different multi-channel real-time PCR systems on the newly optimized assay. Methodology: A triplex real-time PCR assay developed for three key genes encoding virulence and antibiotic resistance in Staphylococcus aureus, namely, lukSF-PV, mecA, and spa was optimized and evaluated using two different real-time PCR instruments (7500SDS and LightCycler 480). Bacterial strains (N=230), including staphylococcal and non-staphylococcal isolates, were Original Research Article Okolie and James; JALSI, 2(4): 145-151, 2015; Article no.JALSI.2015.016 146 used for the study. Results: Following optimization, cycling and data analysis completed within one hour compared with the former three hours. Assay specificity became 100% on both instruments. The negative predictive value (NPV) and the positive predictive value (PPV) rose to 100%. Conclusion: The optimized triplex real-time PCR assay is highly reproducible across the two systems without loss of speed, sensitivity and specificity. The results also suggested that userfamiliarization with each machine operational system would allow assay performance across various real-time PCR platforms currently being used.


INTRODUCTION
Public health challenges from Staphylococcus aureus are associated with molecules which cause therapy refractory conditions.The molecules include staphylococcal proteinA (Spa) encoded by spa gene, penicillin-binding protein 2a encoded by mecA gene, and Panton-Valentine leucocidin (PVL) encoded by phageborne bi-component lukSF-PV operon [1].Spa, coagulase and PVL have been incriminated in diverse S. aureus infections including abscesses, skin and soft tissue infections and musculoskeletal damage [2][3][4], while mecA-driven methicillin resistant S. aureus (MRSA) is of massive diagnostic, therapeutic and infection control importance.Whereas MRSAs first emerged in UK hospital patients [5,6], victims of PVL-associated disease are often young and otherwise healthy persons and outcome is usually independent of hospitalization and staphylococcal methicillin resistance status [7].The convergence of virulence and antibioticresistance signals high-level therapy refractory infections and more significant threat than the two factors separately [8].

Bacteriological
culture-based tests for staphylococcal methicillin resistance take ≥24hrs.One of the diagnostic challenges from the convergence of rapidly mortifying PVLassociated toxicity and methicillin resistance is that there is currently no reliable bacteriological assay for PVL.It has been suggested that application of multiplex PCR to detect multiple genetic markers within the same sample will enhance molecular diagnostics [9].In line with this, several multiplexed assays have been developed in conventional and real-time PCR platforms, some of which have been reported to be ineffectual following evaluation studies.A recent evaluation revealed discrepancies between a triplex real-time PCR assay for nuc, pvl, and mecA [10] and phenotypic antimicrobial susceptibility testing (AST) which led to modification of the same triplex assay into a quadruplex assay [11].To enable application across instruments and to keep performance at par with recent PVL-MRSA assays, we optimized a similar triplex real-time PCR assay for detection of spa, lukSF-PV and mecA [12].Here we report the performance evaluation outcomes using two different multi-channel real-time PCR instruments.

MATERIALS AND METHODS
To enable application across instruments and to keep performance at par with recent assays for MRSA and PVL, the cycling conditions and reaction factors of a triplex real-time PCR assay originally described elsewhere [12] were optimized.Two real-time PCR instruments (ABI 7500SDS: Applied Biosystems, CA, USA and LightCycler™ 480: Roche, Germany) were used for the study.Staphylococcal type culture strains (n=50) used for this study are listed (Table 1).The assay was further evaluated using randomly selected local clinical staphylococcal strains (n = 121) isolated in the Nottingham area during the period August 2003 to December 2004 from blood, sputum, pleural effusion, cerebrospinal fluid (CSF), breast fluid, spinal fluid, wound swab, vaginal swab, abscess, urine and other specimens analysed by the Queen's Medical Centre (QMC) microbiology laboratory.The above 171 strains were collected and stored (-80°C) by Dr. Richard Spence who used them for validation of microarrays [13].Additional S. Aureus strains (n=17) and other staphylococci (n=20) were also used for the evaluation including S. capitis, S. sciuri, S. hemolyticus, S. saprophyticus, S. lugdunensis and S. epidermidis.
Non-staphylococcal bacterial strains/DNA (n=22) including Escherichia coli, Group A Streptococcus, Pseudomonas spp., Proteus spp., and Klebsiella spp.were also used for performance evaluation.Thus, a total of 230 DNA samples from bacteria were used for the evaluation.Extraction of DNA from the bacterial isolates was performed by boiling and centrifugation as described elsewhere [14].
All primers were supplied by Sigma (Sigma Genosys, Cambridge, UK).Probes used on the AB7500 were exactly the same as published by Nakagawa and were purchased from Applied Biosystems (Applied Biosystems, Warrington, UK).The LCR670 probe used for the detection of mecAon the LC480 (Table 2) was purchased from TIB MOLBIOL (TIB MOLBIOL, Berlin, Germany).Manufacturer-specific master-mixes for the two instruments were used.To evaluate the specificity of the optimized assay, sequencing followed by standard nucleotide BLAST were used to confirm the identity of the PCR amplification products.Antibiotic susceptibility test (AST) was performed for staphylococcal oxacillin susceptibility.Tube coagulase (TC) test was used to confirm S. aureus.Ten-fold serial dilutions were performed on 0.5MacFarland standardized inoculum to determine the limit of detection (LoD) of the optimized assay.Bacterial DNA was extracted from each dilution and used for real-time PCR assay.

Statistical Analysis
The R 2 values were calculated from the mean of five most linear points, an approach which increases the accuracy of the DNA quantification by allowing the real-time PCR software to perform a five-point standard curve [15].Sensitivity, specificity, negative predictive value (NPV) and positive predictive value (PPV) of the triplex real-time PCR assay were analysed according to the Clinical and Laboratory Standards Institute (CLSI) Guidelines for molecular diagnostics [16].

RESULTS
The optimized PCR cycling conditions which increased assay speed on both instruments and without spectral cross-talk consists of an initial single cycle for 5 minutes at 95ºC (to activate the polymerase system) and 40 cycles of twotemperature cycling consisting of 15s at 95ºC (for denaturation) and 5s at 60ºC (for polymerization).The optimized triplex real-time PCR assay completed one diagnostic cycle including data analysis within one hour.This is a major speed advantage over the original assay by Nakagawa et al. [12].In addition to the optimized PCR cycling conditions, the oligonucleotide primer-probe sets driving the triplex real-time PCR assay are listed (Table 3).Other optimized PCR factors which necessarily removed spectral cross-talk are also presented (Table 4).When the crossing point (Cp) values were plotted against the log 10 of the bacterial DNA load, a good linearity was observed over the range from 1.0-2.0x10 8 colony forming units/ml (abbreviated: CFU/ml), which was confirmed phenotypically by serial dilution methods.This sensitivity result applied to both instruments.
Using MRSA strains from NARSA to study assay specificity in monoplexes and multiplexes including NRS194, NRS192, and S. aureusUSA400 (MW2) whose genomes harbour all the three key genes, the assay reliably identified spa, lukSF-PV and mecA on both instruments.
PVL-negative MSSA strains generated only the spa signal, PVL-positive MSSA strains yielded positive signals for spa and lukSF-PV, while producing negative signal for mecA.Methicillin-resistant coagulase-negative staphylococci (MRCoNS) including S. epidermidis strain NRS8, S. hemolyticus NRS9 and S. epidermidis NRS69 all gave positive signals for mecA and and not for spa and lukSF-PV.Also the 22 bacterial strains of nonstaphylococcal background generated no amplification signal with the triplex PCR assay.
When we evaluated the triplex real-time PCR assay on direct bacterial cultures from frozen stocks, agar plates, and from 0.5McFarland broth used for phenotypic identification and AST, they all yielded concordant gene detection results.Despite differences in user interface and display of emission signals (Fig. 1), discordant performance characteristics were not found between the two instruments.

DISCUSSION
Optimization of the cycling conditions which led to completion of cycling and data analysis under one hour.This compares favourably with a recent real-time PCR assay reported by Pichon et al. [11].Interestingly, the Pichon assay was developed following evaluation of a similar 2005 triplex real-time PCR assay [10] which the Pichon group found erroneous that they modified and upgraded it to a quadruplex assay.The reduction in assay cycling time from 3 hours to less than one hour will empower microbiology laboratories to provide speedier diagnostic service.This will help reduce the rapid victim deterioration associated with therapy-refractory S. aureus PVL syndrome.Sensitivity and good linearity over the range 1.0-2.0x108 CFU/ml suggests the optimized triplex real-time PCR assay is of comparable sensitivity as the original assay developed by Nakagawa using a different machine (AB7900HT).Therefore the gain in speed associated with the newly optimized assay was without loss of sensitivity.PCR was commended by early experimenters for its sensitivity [17].Also, the fact that no amplification signal was detected from the 22 non-staphylococcal bacteria examined with the optimized triplex real-time PCR assay shows the assay did not lose specificity (precision and accuracy) following optimization.The newly optimized triplex real-time PCR assay identified all mecA-positive oxacillin resistant strains, including high-level (MIC oxacilin ≥32 g/L) and very low-level (0.5 g/L) expressors.The assay also detected mecA gene from all the SCCmec types collected by Dr Spence, including those local to Nottingham as well as the National (NCTC and NCIMB) and international (NARSA) type culture strains.Thus, the capacity of the optimized real-time PCR assay to detect the expected gene sequences as confirmed by sequencing and BLAST was 100%.This represents a 100% positive predictive value (PPV=100%).Evaluation of a similar assay developed in 2005 [10] showed it was unable to detect some SCCmec types which encoded phenotypically confirmed oxacillin resistance [11].Also, the 22 eubacteria of nonstaphylococcal background yielded no amplification signal, representing a negative predictive value (NPV) of 100%.Other than obvious differences in the user interface and signal output style (Fig. 1), there was no difference in speed between the two instruments.

CONCLUSION
NPV and PPV values of 100% suggest that the triplex real-time PCR assay is capable of confirming or ruling out the presence of the three key genes in a given bacterial isolate.Thus, the primer-probe sets for lukSF-PV, mecA and spa yielded 100% gene detection specificity on both instruments.The sensitivity, speed, specificity and reproducibility generated from using the triplex real-time PCR assay on both instruments show the assay is readily adaptable for use in a routine diagnostic microbiology laboratory across different instruments.The sensitivity and specificity of the assay on direct detection from broths and colony PCR suggests the applicability of the assay for direct detection of the three key genes from biological samples.Completion of cycling and data analysis under an hour means the optimized assay is speedy enough to support timely infection control and clinical therapeutic interventions than would be possible by conventional microbiological culture-based assays.
Despite difference in signal detection emission channels, the triplex real-time PCR assay developed using one machine had been packaged and operated on another real-time PCR proprietary instrument(s) from a completely different manufacturer.What is required is sound understanding of the operational style of each unique instrument, including the wavelengths for the different emission channels as well as the output signal and the user-interface data presentation style.

Table 1 . Reference staphylococcal strains (n = 50) and their sources
a Identity of isolate at source, b NARSA, Network for Antimicrobial resistance in Staphylococcus aureus c NCTC, National Collection of Type Cultures, d NCIMB, National Collection of Industrial and Marine Bacteria