The extraction efficiency of manual, semi-automated, and fully automated systems is analyzed.
By Ira Ashman
Prior to the advent of polymerase chain reaction (PCR) chemistry in 1983, detection of infectious agents in biological specimens relied heavily on labor-intensive and often flawed culture-based methods. PCR provided researchers and clinical lab professionals a quicker and easier option to detect infectious agents, as results became available in a matter of hours. Yet original PCR formats proved less than ideal for most clinical lab applications, notably for workflow complexity and multiple instrument requirements together with high potential for errors resulting from cross-contamination of samples with amplicons.
The 1990s saw a major leap forward with the development of real-time PCR, which employed fluorescence detection technology and simplified several aspects of the workflow. Varying degrees of automation evolved that not only provided workflow advantages and permitted use of closed systems to essentially eliminate cross-contamination errors but still required multiple instruments and workstations to perform testing. Several target and label amplification-based technology alternatives came into play over the years, yet real-time PCR emerged as the dominant clinical lab modality for infectious disease detection in terms of both commercial kits and user-developed protocols.
|Figure 1. Click to enlarge.|
Several systems that offer varying levels of automation and modularity are currently available on the market, including the first fully automated system capable of performing manufacturer-developed tests and user-defined protocols (UDPs). The efficiency and consistency with which a particular approach can extract nucleic acids from microorganisms in biological samples is a critically important contributor to assay sensitivity, the maximization of which is, in turn, critical to the medical utility of assays.
This article presents results from internal feasibility studies that compare the extraction efficiencies of three systems with the following features:
The manual extraction and purification method for DNA uses either fast-spin columns or vacuum procedures and for this study, purified nucleic acids were analyzed using a stand-alone PCR instrument. The manual protocol starts with enzymatic lysis of bacteria and capture of nucleic acids on silica gel membranes. Buffers and enzymes are optimized to lyse samples, stabilize nucleic acids, and enhance selective adsorption to the membranes. Lysed samples are then treated with alcohol and loaded onto the spin columns. Two wash steps remove trace quantities of any remaining potential PCR inhibitors, and the purified DNA is eluted either with water or a low-salt buffer.
|Figure 2. Click to enlarge.|
The automated nucleic acid isolation and purification system is capable of processing up to 32 samples per batch. Purified nucleic acid samples can be transferred to a separate PCR instrument for amplification and detection. A combination of lysis buffer and proteinase K provide for cell disruption and protein digestion. Nucleic acids bind to the surface of magnetic glass particles, which are sequestered and washed to remove cellular debris, and nucleic acids are eluted from the beads at high temperature. The instrument transfers reagents from container strips into a processing cartridge, after which lysates are transferred from the sample cartridge into reagent wells. The reaction mixture is then moved to a second row of wells, which contain additional reagents. Next, the instrument sequesters beads in order to remove supernatants, and transfers the beads to a cartridge with wells that contain an elution buffer. Finally, purified nucleic acid samples are transferred to a storage cartridge for subsequent analysis.
The BD MAX system integrates sample lysis, extraction, purification, amplification, and detection in a fully-automated walk-away system capable of running 24 samples at a time.1 It uses a combination of lytic enzymes and extraction reagents to generate samples for subsequent amplification and detection. Lysis liberates nucleic acids, which are captured on high-affinity magnetic beads. The instrument employs a magnet to sequester the beads for removal of the supernatant. The first reagent solution is added, beads are washed and incubated with a second reagent solution, and the supernatant containing target for analysis is mixed with a third reagent in preparation for PCR analysis. All reagents and consumables required for lysis, nucleic acid extraction, purification, amplification and detection are loaded into a Unitized Reagent Strip (URS). Strips are loaded into the instrument, which runs through the entire assay process in walk-away mode.
|Figure 3. Click to enlarge.|
In real-time PCR, a useful metric for measuring the relative performance of an assay is percent efficiency, which is defined by the following equation: E = (10-1/slope – 1) x 100. The slope is derived from a graph in which the x-axis contains the log of analyte concentration, and the y-axis contains Ct, the cycle threshold for each data point. Efficiency levels between 90% and 110% represent the industry standard for efficacious real-time PCR analysis.2
Results from two sets of feasibility experiments clarify performance differences among the three aforementioned systems. A first set of experiments compares the automated extraction system (which requires a separate instrument for PCR amplification/detection) and the fully automated system (which performs all steps on a single instrument) for efficiency in extracting DNA from a bacterial target. To this end, it was necessary to first assess the efficiency of the PCR portion of each system. A sample of Salmonella typhimurium spiked to a level of 107 CFU/ml was extracted in both systems. Eluted DNA was subject to five ten-fold dilutions, and each of the five diluted samples, which provided from 105 down to 101 copies of DNA per PCR reaction mixture, were amplified on the respective instruments. The results (Figure 1) show that for this assay, the PCR portions of both systems are inherently efficient (automated extraction/PCR: E = 97.2%; full automation: E = 93.9%).
To compare extraction efficiencies of both systems (i.e., automated extraction followed by PCR and the fully automated system), a quantitated S. typhimurium suspension was subject to five ten-fold dilutions prior to extraction, with each of the five diluted samples run in triplicate through each system. Results (Figure 2) show that the PCR profile for the fully automated system fits within optimal limits (E = 105.9%), whereas the PCR profile for the automated extraction is outside these limits (E = 70.1%). Extraction on a fully automated system also exhibited higher sensitivity than the automated extraction system, since it was able to detect analyte in the most dilute sample.
A second set of experiments sought to compare all three systems (i.e., manual extraction/PCR, automated extraction/PCR, and full automation) in order to determine how extraction-related differences affect the potential sensitivity of detection. To mimic actual clinical samples, a suspension of S. typhimurium was diluted with buffer containing a stool matrix to the level of 5 x 106 CFU/ml. From that, four additional ten-fold dilutions were made using the same buffer/matrix combination. The five resulting solutions were then split three ways. Two sets were run through the automated and manual extraction procedures respectively and the resulting DNA was analyzed using a PCR instrument. The third sample set was extracted and amplified in a single operation using the fully automated system.
|Figure 4. Click to enlarge.|
Comparisons of extraction efficiencies for the manual method versus the fully automated system and the automated extraction versus the fully automated system are shown in Figure 3 and Figure 4, respectively. The sample with the highest analyte level (5 x 106 CFU/ml) yielded Ct values of 20.73 for the fully automated system; 21.38 for automated extraction; and 20.96 for the manual method. As the target levels decreased, the degree of divergence increased between the three systems. Although each of the three systems detected the target at the lowest level tested, the fully automated system did so at least four cycles earlier in each instance.
The findings of this study suggest that the fully automated BD MAX system may have equivalent and/or better extraction efficiency compared to an automated extraction system or a manual extraction method and this may result in a higher potential for PCR amplification and detection on a single platform, particularly for samples with low target loads.
1. BD Diagnostics, 2012, BD MAX System User’s Manual (8088453), Revision 06, BD Diagnostics, Sparks, MD.
2. Rebrikov, D.V and D. Yu. Trofimov, “Real-Time PCR: A Review of Approaches to Data Analysis,” Applied Biochemistry and Microbiology, 42 (2006): 455–463.
Ira Ashman, BS, is a scientist at BD Diagnostics, Sparks, MD. He can be reached at email@example.com.