A target-dependent activation step is combined with isothermal amplification, helicase-dependant amplification, and DNA array-based detection on modified silicon chips to create a visually detectable signal.
Appropriate early treatment of hospital-acquired infections (HAI) has been associated with lower treatment costs and improved patient outcomes. Current culture-based methods for diagnosing HAI are either slow or poorly sensitive/specific; therefore, broad-spectrum antimicrobial therapy is often initiated when there is clinical suspicion of infection. This approach may not be effective when dealing with inherently resistant organisms and can result in iatrogenic infections. Amplification and detection of nucleic acids (molecular diagnostics) has improved diagnosis for many indications, and sample-to-result technology is of great demand in the clinical diagnostics laboratory, improving workflow and efficiency.
Early sample-to-result platforms were highly complex instruments in a centralized laboratory that required specialized technicians to operate, such as the TIGRIS system and ProbeTec systems used for STD pathogen screening and blood screening. To optimize efficiency and cost, these platforms perform tests in large batches, resulting in slower turnaround time for individual samples. An additional limitation for these approaches is the inability to multiplex, limiting information and increasing cost.
Cepheid has pioneered a user-friendly, multiplex-capable instrument, GeneXpert, using real-time PCR technology. However, limited multiplexing capabilities still do not allow for information-rich, panel-based testing, and high expense is a burden on hospital laboratory budgets. Nanosphere has addressed multiplexing by coupling target or signal amplification with DNA probe array–based technology in the Verigene SP system; however, its approach requires multiple instruments and disposables. The complexity and slower speed of either of these approaches make them unacceptable for the point-of-care setting.
To address requirements for improved turnaround time, better ease-of-use, and lower cost, Great Basin has developed a molecular diagnostics approach combining a novel target-dependent activation step with the isothermal amplification method, helicase-dependant amplification (HDA), and DNA array-based detection on modified silicon chips to create a visually detectable signal. This approach is executed in a sample-to-result format using a single low-cost disposable run on an electromechanically simple instrument. Recently, a test developed to detect toxigenic Clostridium difficile on this platform received 510(k) clearance by FDA. Herein, we describe application of this technology to the detection of staphylococcal species and the methicillin-resistance determinant mecA directly from positive blood cultures.
PCR technology has demonstrated excellence for amplification of multiple DNA target sequences. However, the technology is limited by the speed of thermal cycling by the instruments. Isothermal amplification methods have appeal due to the simplicity of automation; using a heater at a single temperature makes thermal management much easier. Additionally, because the speed of amplification is limited only by the speed of the enzymes, turnaround time may be improved. However, none of the described isothermal methods, which include helicase-dependant amplification (HDA), loop-mediated amplification (LAMP), or recombinase/polymerase amplification (RPA) have any capability to prevent primer-primer interactions that lead to amplified primer artifacts, which limit sensitivity and multiplexing.
To overcome this problem we developed a novel method called blocked-primer helicase-dependant amplification (bpHDA),
|Figure 1: BpHDA. At lower temperatures, blocked primer (black) bound to the target sequence (white) cannot be extended by DNA polymerase due to the presence of a 3’ blocking group (X = 3-carbon spacer). At elevated temperatures, RNase H2 is activated, and a ribonucleotide (r) near the 3’-end of the primer is cleaved, liberating a 3’-hydroxyl group, allowing for primer extension by DNA polymerase (circle).|
which uses the isothermal nucleic acid amplification method HDA to exponentially amplify target DNA sequences coupled with a blocked primer/RNase H2-mediated target-specific “hot start.” In bpHDA, modified blocked primers are constructed with a single ribonucleotide linkage 3 to 5 bases upstream of a 3’-end block, to prevent primer extension. Once blocked, primers hybridize to complementary target sequences, and thermostable RNase H2 derived from Pyrococcus abyssi is activated, cleaving the ribonucleotide linkage in the primer present in duplex DNA. The short segment of the primer 3’ of the ribonucleotide dissociates, liberating the block and creating a free 3’-hydroxyl, which is now capable of primer extension. RNase H2 used here has very little activity at temperatures below 40ºC and is highly active at 65ºC, the temperature required for HDA to amplify target sequences optimally. Because primer-primer hybrids are unstable at elevated temperatures, primer artifact is not amplified. Unblocked primers can participate in DNA amplification reactions using HDA, which uses the enzyme DNA helicase to unwind double-stranded amplified DNA unlike heat in PCR. Exponential amplification occurs identically to PCR with primers being extended by DNA polymerase (Figure 1).
Because primer artifact is mitigated in bpHDA, we are able to use higher primer concentrations allowing for faster reactions. For example, the genome of Mycobacterium tuberculosis is very GC-rich, making amplification of target sequences difficult due to artifact formation. In an application directed at detecting mutations in the rpoB gene of M. tuberculosis, we compared the performance of bpHDA to HDA with unmodified primers. In standard HDA, we were only able to use up to 50 nM primer concentrations. Under these conditions amplification reactions were slow, requiring one hour to generate a detectable signal from 30,000 copies of genomic DNA with a limit of detection of 100 copies even after 90 minutes of amplification time. For bpHDA, similar results were observed at 50 nM primer amounts; at higher primer concentrations, a single copy of genomic DNA was detectable after 40 minutes of amplification time, whereas with standard HDA, only primer artifact was observed.1 For other bacterial genomes with more typical GC content, rates of amplification are very high. For example, amplification of the tcdB gene within C. difficile genome occurs with a doubling time measured to be 28 seconds, allowing for detection of the amplification of a single copy of genomic DNA in 17 minutes by real-time amplification reactions.2 An additional benefit of bpHDA is improved consistency in amplification performance compared with standard HDA protocols. In a duplex designed to co-amplify toxigenic Clostridium difficile and a sample process control organism, Staphylococcus aureus, we observed robust and consistent amplification with bpHDA, whereas standard HDA yielded inconsistent amplification due to the presence of competing primer artifact.2 This is particularly significant at lower-input organism amounts.
Array technology currently is dominated by fluorescence or electrochemical signaling methods. However, these technologies require either expensive imaging approaches or more complex surface preparation. The Portrait PA5000 system utilizes inexpensive silicon chips modified such that intermolecular interactions that occur on the surface of the chip are amplified and can transduce intensity changes in the surface color that are apparent to the naked eye, permitting visual detection of attomole quantities of nucleic acid hybrids. Because of this, low cost CMOS or CCD cameras can be used to capture chip images for subsequent image analysis, lowering device cost.
|Figure 2. Portrait detection approach. Once bpHDA amplification is complete, biotin-labeled amplicons are hybridized to target-specific probes immobilized on a modified silicon surface. After an optional wash step, anti-biotin antibody/HRP conjugate is incubated on the chip surface. After a required wash, the surface is incubated with a precipitable formulation of the substrate, TMB. After TMB incubation, the surface displays a visibly detectable signal, which can be imaged using CCD or CMOS cameras.|
Epitaxial silicon wafers are coated and processed using semiconductor processing equipment, allowing for highly scalable, low-cost production. The wafer coating consists of hydrophobic polymeric siloxane that is conformal to the molecularly flat polished silicon surface. The molecular flatness and low chemical reactivity of the modified silicon surface combine to exhibit very low nonspecific binding properties. To facilitate probe attachment, two additional chemical coatings are applied to create a surface with functional aldehyde groups. We then apply DNA probes modified with terminal hydrazide groups, creating a stable hydrazone linkage allowing for one-step chemical attachment. The chemistry is very efficient, allowing for a low-cost process, and stability of probes on the surface has been measured at greater than one year with no loss in activity.
The tests are configured to perform in an ELISA format using reagents that are well-characterized commercially and are very stable (Figure 2). To facilitate detection, bpHDA reactions use biotin-labeled primers. Biotin-labeled DNA targets are hybridized to probes arrayed on the chip surface. To detect target sequences, the array is exposed to an anti-biotin antibody/HRP conjugate, washed, and then incubated with a precipitable formulation of the substrate TMB. Enzymatic product deposits on the chip surface and is observed as a blue-gray to sky-blue color. The LLOD is approximately 0.3 pM in this format for the detection of DNA duplexes. Because the quantity of bpHDA amplified products (10-100 nM) is orders of magnitude above the limit of detection, short assay times can be used (5 minutes hybridize, 3 minutes conjugate, 2 minutes TMB). Combined with rapid amplification, this assay approach has the potential for very rapid diagnostic test results. This is particularly significant for test-and-treat indications, where the physician would like to make a decision before the patient leaves the office. For example, treatment of influenza virus is very time-sensitive and requires rapid results. Current lateral flow antigen-based tests satisfy the requirements for rapid tests; however, they suffer from relatively poor sensitivity.3
The system described herein, termed Portrait PA5000, is a sample-to-result test platform (Figure 3). It is designed to require no user maintenance and to be easy to operate through a simple user interface. Liquid clinical specimens are loaded into a self-contained single-use cartridge that contains all of the reagents necessary and collects all waste. The customer then enters patient information and initiates a test routine performed by the instrument; no additional hands-on steps are required.
The cartridge is designed as a mesofluidic device, processing relatively large microliter volumes to avoid the physical effects of fluids that can be difficult to manage in microfluidic devices. The cartridge contains all reagents necessary to perform the assay in blister packs except for the amplification reagents, which are stored as a lyophilized pellet. The base of the injection-molded disposable cartridge includes features such as reaction chambers, a waste chamber, and channels to direct the movement of fluids. Integrated into the base cartridge are lance devices for the reagent blister packs, stirring devices, and lyophilized reagents. Reaction chambers and fluid channels are covered with a clear layer of thermoplastic that is laser-welded to the base cartridge to form liquid-tight features. The reagent blister packs, thermal transfer pads, and hydrophobic/oleophobic vents are then attached to complete the cartridge.
|Figure 3. Portrait cartridge and PA5000 analyzer|
To execute the biochemistry, there are two chambers for sample prep, an amplification chamber, a detection chamber, and a waste chamber that collects all liquid used in the assay. Three of the blister packs are available for sample prep. The device is very flexible for sample prep because of the presence of magnetizing and mixing capabilities, permitting multiple extraction approaches including enzymatic, heat, and shearing as well as dilution, filtration, or magnetic particle approaches to cell or nucleic acid capture. The amplification chamber is designed for precise temperature control by using conduction heaters located above and below the chamber. The cartridge design includes thermal pads to optimize thermal conductivity. The detection chip consists of a ~7-mm2 epitaxial silicon chip with spotted capture probes and is bonded into the cartridge detection chamber. Up to 8 × 8 arrays of nucleic acid probes can be applied to the chip surface using a noncontact print head.
The instrument is controlled by a low-cost PC-based laptop computer. Great Basin’s proprietary Portrait PA5000 analyzer application is installed on the laptop and controls the device through a single USB 2.0 connection. To minimize laboratory space requirements, the instrument dimensions are 6.5 × 21 × 17 in. The Portrait PA5000 analyzer comprises three primary subsystems: reagent flow control (valve motor drives, optical sensor), thermal control, and the optical imaging system. Once the operator inserts the cartridge into the Portrait PA5000 analyzer, enters required patient information, and initiates the assay sequence, the device performs the extraction sample dilution, amplification, and detect step automatically and displays the result on the user interface and generates a report.
|Figure 4. Real-time amplification comparison of bpHDA with standard HDA. Panel A: Real-time amplification curves for 100 copies of MRSA genomic DNA. Green lines indicate standard HDA curves. Red lines indicate bpHDA curves. Panel B: Melt peaks for amplified product. Product Tm = 78.5°C. Gold lines indicate standard HDA melt profile. Blue lines indicate bpHDA melt profile.|
To execute an assay, a loaded cartridge is placed into an opened door. Once the door is closed, the cartridge is motivated into the instrument, where it is secured on a platform allowing access to the cartridge from above and below. Beneath the cartridge reside valve actuators, which can serve to control fluidic movement, and lance actuators, which lance blister packs. Above the cartridge reside mix motors and stepper motors, which can flatten blister packs and drive reagent into fluidic channels. The mesofluidic movements are achieved by compressing the blister packs or pressing on the diaphragm of the processing chambers with a motor and foot assembly. Compression valves are used to isolate regions of the cartridge for mixing, amplification, and detection. Optical sensors located adjacent to the fluid paths in the cartridge are used to determine fluid movement and endpoints for chamber-filling activities. Multiple resistive heaters with thermocouple feedback are used to control temperature for extraction, amplification, and hybridization processes. Active stirring of the solutions is achieved by using magnetically coupled stir bars in the appropriate chambers. For isothermal DNA amplification, the chamber is fluidically isolated and maintained at 65° ±2°C by direct contact with a controlled heat source. For detection, the amplified sample is diluted with hybridization buffer and introduced into a chamber where the modified silicon chip is affixed. As with prior steps, fluidic movements and temperature are controlled for the hybridization, washing, and signal development steps. The resulting visible features are captured by a digital camera. Processing and filtering techniques define true reacted elements. Multiple custom algorithms query pixel intensity and intensity gradient directionality to determine presence or absence of a signal on each array feature. Once the optical reader software has determined the presence or absence of signal on each array feature, a call logic tree is used to determine the assay result, which is reported automatically.
Nosocomial infections are an important cause of morbidity and mortality, with 2 million diagnosed cases in the United States each year causing 90,000 deaths.4 Staphylococcal species are the leading cause of nosocomial infection worldwide, and up to 60% of all staphylococcal infections are methicillin resistant. Staphylococcus aureus is widely recognized as a significant pathogen, but the increased use of indwelling devices has implicated coagulase-negative staphylococci in an increased risk of morbidity and mortality, as well.5, 6 Appropriate early treatment of bloodstream infections has been associated with improved patient outcomes, and clinical-outcome studies have shown that reducing the time to diagnosis decreases the patient’s length of stay and rates of morbidity and mortality.7, 8, 9 Recently described real-time PCR-based formats for MRSA/S. aureus identification have reduced the time of diagnosis from 24 to 48 hours using standard microbiological approaches to 1 to 2 hours, shortened length of stay, and reduced treatment costs because of more-rapid initiation of appropriate antibiotic treatment.10 However, these tests do not provide information about coagulase-negative staphylococci, with their presence implied only by a negative test result for gram-positive cocci in clusters of positive blood culture. Because up to 90% of coagulase-negative staphylococci are methicillin-resistant, clinicians typically treat suspected blood stream infections with vancomycin. However, several reports have shown that vancomycin was less effective than beta-lactam antibiotics for treatment of S. aureus infections and its misuse could potentially lead to resistance. Rapid determination of mecA gene status for coagulase-negative staphylococci may initiate switching of methicillin-sensitive infection treatment from vancomycin to beta-lactam drugs; early clinical studies have observed a positive effect on patient outcome.11
Appropriate management of infection also includes determining when treatment is not required. It has been reported that contaminating bacteria account for 2.5% of all blood cultures on average, or roughly 25% of all positive blood cultures, with 70 to 80% of these cases being caused by coagulase negative Staphylococcus.12 Treatment of these cases may be unnecessary and lead to longer lengths of stay and higher treatment costs.13 In order to distinguish contaminants, one method of choice is to serially draw blood, typically two to four times. When cultures are positive from multiple blood draws, the positive predictive value (PPV) for bacteremia improves due to the diminishing probabilities that the same organism is present in each bottle by random chance.14 It has been proposed that identity of true infection can be improved by determining species relatedness between blood culture bottles; if the organism in each bottle is identical, it is highly likely to be a true infection. However, phenotypic identification methods require subculturing, adding significant time to the process of identifying true infections. Additionally phenotypic methods are not highly accurate, and are, therefore, not likely to aid in identifying true infections for species that are difficult to reliably identify. Strain relatedness can be determined by RFLP; however, this requires subculturing and is a labor-intensive method that requires sophisticated analysis.
To address the above unmet need, we have developed Staph ID/R, which identifies the major pathogenic staphylococci to the species level and detects the presence of the mecA gene, which imparts methicillin resistance. A multiplex bpHDA reaction is performed and the resultant amplicon is hybridized to an array of mecA gene and species-specific probes. Probes are printed that detect the seven most prevalent staphylococci to the species level representing >98.5% of all positive staphylococcal blood cultures, with the remaining species detected by probes that cover groups of species. In a first version of the application, we amplified staphylococcal positive blood cultures using standard HDA. The data demonstrated excellent clinical sensitivity and specificity, suggesting the approach is a potentially useful tool for patient management.15 We subsequently have adapted the amplification reaction to bpHDA, speeding amplification and improving assay consistency, and have integrated the described test into the Portrait PA5000 analyzer. The bpHDA reaction for Staph ID/R requires the amplification of both the tuf gene and mecA gene within staphylococci. As shown in figure 4 for the amplification of MRSA blood cultures diluted into negative blood culture, within 10 minutes of bpHDA amplification using real-time methods one can detect 100 copies of MRSA genomic DNA, whereas it takes 20 to 23 minutes to detect using standard HDA. In model experiments we have determined a doubling time of 20 seconds for this assay allowing for the amplification of a single copy of genomic DNA to detectable levels of amplicon within 15 minutes. An additional benefit, as is seen in Figure 4, is improved consistency and quality of the amplification. For HDA amplification products, clean melt peaks are observed indicative of amplification of the desired product; with standard HDA, significant artifact formation is observed in the melt peaks. This effect is pronounced at lower input copies of target.
|Figure 5. Representative results for the detection of bacterial species using Staph ID/R. Blood cultures positive for the indicated bacterial species were tested using the Staph ID/R protocol, and CCD images of resulting chips are shown. Probe map is shown in right panel. F = fiducial marker, DC = detect control, HC = hybridization control, SGe = Staphylococcus genus, SA = S. aureus, SE = S. epidermidis, SHa = S. haemolyticus, Sca = S. capitis, Slu = S. lugdunensis, SHo = S. hominis, SW = S. warneri, EFa = E. faecalis, mec = mecA.
To perform the Staph ID/R test, the operator simply pipettes positive blood culture media directly into the cartridge. After closing the sample port, the cartridge is inserted into the analyzer, sample information is entered, and the test is initiated using a graphical user interface. At the conclusion of the test, results are displayed showing what staphylococcal species is present and whether or not the mecA gene was detected. Sample results from instrument runs are shown in Figure 5. Similar levels of sensitivity and specificity are observed in the instrument as on the bench. For example, as the test is designed there is a single base difference between S. aureus and S. warneri. As is seen in Figure 5, excellent discrimination is observed between these two species, with no cross-reactivity observed. Additionally, sensitivity is good for all species detected including two species (S. warneri and S. lugdunensis) that amplify more slowly at lower input amounts of target DNA because of mutations under the primer sequence. As is seen in Figure 5, assay results are similar for blood cultures tested directly (108 CFU/mL) or for those diluted 1/100 into negative blood culture (108 CFU/mL). This is significant as the LOD necessary for detection of positive blood cultures requires this range of performance.16
We have plans to initiate clinical validation of this test in planned studies upcoming on both of the major blood culture instrument platforms, the BACTEC system (Becton Dickinson) and the BacTAlert system (BioMérieux).
Currently Great Basin is developing a suite of products designed to detect pathogens present in positive blood culture bottles with plans to migrate some of these tests to detection of pathogens directly from the blood stream. We have successfully developed a panel that can identify seven clinically significant Candida species directly from positive blood cultures and we also are developing broader multiplex panels for gram positive and gram negative organisms along with drug-resistance-gene detection. Further, expansion of the content of our FDA-approved toxigenic C. difficille assay to detect hypervirulent strains that are associated with increased rate of relapse of infection is in development; identity of multiple ribotypes, not just 027, will be included.
Currently a single cartridge processing instrument is available. A multiple cartridge processor is also in development. As another approach to increase throughput and lower cost per test, we are developing a cartridge that can process two samples at a time, leveraging the same blister packs. Also, Great Basin has developed a test for detecting rpoB gene mutation in Mycobacterium tuberculosis on a further simplified device that is consistent with the ASSURED goals for developing world diagnostic test requirements.1
In the future, there is a significant market opportunity for CLIA-waived tests to be performed directly in the physician’s office (POL) setting. This includes rapid test-and-treat indications such as viral and bacterial respiratory pathogens as well as women’s health indications such as detection of pathogens that cause bacterial vaginosis and vaginitis and sexually transmitted disease pathogens, including Chlamydia trachomitis and Neiserria gonorrhea. The cost/speed profile of our test platform makes it attractive for the point-of-care setting, where highly specific and sensitive rapid detection can direct appropriate therapy before the patient leaves the office.
1. Ao W, Aldous S, Woodruff E, et al., “Rapid detection of rpoB gene mutations conferring rifampin reistance in Mycobacterium tuberculosis,” Journal of Clinical Microbiology, 50: 2433-2440, 2012.
2. Hicke B, Pasko C, Groves B, et al., “Automated detection of toxigenic Clostridium difficile in clinical samples: Isothermal tcdB amplification coupled to array-based detection,” Journal of Clinical Microbiology, 50: 2681-2687, 2012.
3. Ginocchio CC, Zhang F, Manji R, et al., “Evaluation of multiple test methods for the detection of the novel 2009 influenza A (H1N1) during the New York City outbreak,” Journal of Clinical Virology 45: 191-195, 2009.
4. Centers for Disease Control and Prevention. “National Nosocomial Infections Surveillance (NNIS) system report, data summary January 1992-June 2001, issued August 2001,” American Journal of Infection Control, 29: 404-421, 2001.
5. Gastmeier P, Geffers C, Sohr D, et al., “Surveillance of nosocomial infections in intensive care units: current data and interpretations,” Wien Klin Wochenschr, 115: 99-103, 2003.
6. Silva HL, Strabelli TM, Cuhna ER, et al., “Nosocomial coagulase-negative staphylococci bacteremia: five year prospective data collection,” Journal of Infectious Diseases, 4: 271-274, 2000.
7. Shorr AF, Micek ST, et al., “Inappropriate therapy for methicillin-resistant Staphylococcus aureus: resource utilization and cost implications,” Critical Care Medicine, 36: 2335-2340, 2008.
8. Ibrahim EH, Sherman G, Ward S, et al., “The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting,” Chest, 118: 146-155, 2000.
9. Pittet D, Tarar D, and Wenzel R P, “Nosocomial bloodstream infection in critically ill patients: excess length of stay, extra costs, and attributable mortality,” JAMA, 271: 1598-1601, 1999.
10. Stamper PD, Cai M, Howard T, et al., “Clinical validation of molecular BD GeneOhm StaphSR assay for direct detection of Staphylococcus aureus and methicillin-resistant Staphylococcus aureus in positive blood cultures,” Journal of Clinical Microbiology, 45: 2191-2196, 2007.
11. Schweizer ML, Furuno JP, Harris AD, et al., “Comparative effectiveness of nafcillin or cefazolin versus vancomycin in methicillin-susceptible Staphylococcus aureus bacteremia,” BMC Infectious Diseases, 11: 279, 2011.
12. Hall KK and Lyman JA, “Updated review of blood culture contamination,” Clinical Microbiology Reviews, 19: 788-802, 2006.
13. Forrest GN, Mehta S, Weekes E, et al., “Impact of rapid in situ hybridization on coagulase-negative staphylococci positive blood cultures,” Journal of Antimicrobial Chemotherapy, 58: 154-158, 2006.
14. Weinstein M, “Blood culture contamination: persisting problems and partial progress,”Journal of Clinical Microbiology, 41: 2275-2278, 2003.
15. Pasko C, Hicke B, Dunn J, et al., “Staph ID/R: a rapid method for determining Staphylococcus species identity and detecting the mecA gene directly from positive blood culture,” Journal of Clinical Microbiology, 50: 810-817, 2012.
16. Marlowe EM, Gibson L, Hogan J, et al., “Conventional and molecular methods for verification of results obtained with BacT/Alert nonevent blood culture bottles,” Journal of Clinical Microbiology, 41: 1266-1269, 2003.
Larry Rea is senior vice president, operations, at Great Basin Corp., West Valley City, Utah.
Brian Hicke is director, research, at Great Basin Corp., West Valley City, Utah.
Wes Lindsey is director, product development, at Great Basin Corp., West Valley City, Utah.
Michael McMahon is senior hardware systems engineer at Great Basin Corp., West Valley City, Utah.
Charles Owen is director, engineering, at Great Basin Corp., West Valley City, Utah.
Robert Jenison is chief technology officer at Great Basin Corp., West Valley City, Utah. He can be reached at firstname.lastname@example.org.