| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Molecular Diagnostics and Genetics |
1 Labor Becker, Olgemöller und Kollegen, Führichstrasse 70, 81671 München, Germany.
aAuthor for correspondence. Fax 49-89-450917-300; e-mail sburggraf{at}labor-bo.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Single-stranded oligonucleotides, which contain little more than primer and probe binding sites, were used as internal controls in real-time PCR assays. Mismatches were included in the probe-binding region of the internal control oligonucleotide (ICO) to prevent probe–control hybridization during the fluorescence acquisition step of the PCR. Amplified ICOs were detected by melting point analysis. ICOs could be added directly to the sample material before DNA extraction.
Results: To demonstrate the feasibility of the new approach, we designed ICOs for the LightCycler hybridization probe assays for Mycobacterium tuberculosis complex, hepatitis B virus, herpes simplex virus, and varicella zoster virus. In each case, the controls did not interfere with detection of the pathogen, but were clearly detectable during a subsequent melting point analysis.
Conclusions: A single-stranded oligonucleotide that mimics the target region of the pathogen but is clearly distinguishable from the target during melting point analysis can serve as a simple, cost-effective internal control for real-time amplification assays. Such control oligonucleotides are easy to design and inexpensive. A costly second probe system is not necessary. Moreover, the internally controlled assay uses only one fluorescence detection channel of the instrument, leaving the second channel free for multiplex applications.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This is an important assay requirement because there are many potential sources of amplification inhibition. For example, if they are not removed during nucleic acid purification, many substances present in clinical samples (e.g., hemoglobin) can inhibit amplification (1)(2). In addition, certain compounds used to extract nucleic acids from sample material, e.g., ethanol or detergents such as sodium dodecyl sulfate, are potent amplification inhibitors.
A simple way to detect inhibitors is to mix a positive control nucleic acid with the sample nucleic acid after sample purification (3). However, because such external controls must be added to each sample in a separate reaction vessel, the costs of the assay are increased and the assay set-up is complicated, especially when the assay involves many samples. Moreover, external controls cannot reveal another possible cause of false-negative results, i.e., inefficient extraction of the target nucleic acid.
Internal controls can overcome many of the limitations of external controls. The most commonly used internal control for PCR is a plasmid that contains a sequence similar to, but not identical to, that of the assay target. The control plasmid is added directly to the crude sample material, and after nucleic acid extraction, it and the target are coamplified with the same set of primers. Because of the slight difference in sequence between control and target, the two amplification products can be differentiated. For example, in real-time amplification procedures, two differently labeled probes recognize the target and the control products in the same reaction vessel (4). This type of internal control is thought to be the most accurate way to control important steps involved in diagnostic amplification protocols (5)(6)(7).
Here we describe a novel approach for the design and use of internal controls that can accurately detect false negatives but do not require time-consuming plasmid construction or a separate detection probe. We also demonstrate its application to LightCycler (Roche Molecular Biochemicals) real-time hybridization probe assays for the Mycobacterium tuberculosis (MTB) 1 complex (MTBC; M. tuberculosis, M. africanum, M. microti, and M. bovis), hepatitis B virus (HBV), herpes simplex virus (HSV), and varicella zoster virus (VZV).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. All oligonucleotides were synthesized by Metabion (Planegg-Martinsried, Germany).
|
internal control oligonucleotides
Internal control oligonucleotides (ICOs) were designed to basically consist of the primer- and probe-binding regions from the target DNA. Mismatches were introduced into the binding site for the detection probe to reduce the melting temperature of detection probe–ICO hybrids. The sequences of the ICOs used in this study are given in Table 1
. ICOs were synthesized by Metabion. (Note: To avoid contamination it is important not to order ICOs at the same time that you order primers and probes from the same company.)
In a room separate from the area in which samples were prepared and the PCR was set up, the oligonucleotides were dissolved in Tris-EDTA buffer (1 mmol/L Tris-HCl, pH 8.0; 0.01 mmol/L EDTA, pH 8.0). ICOs were diluted to 5000 copies/µL in Tris-EDTA and stored in aliquots at –70 °C. Thawed once, ICOs were stored at 4 °C up to 5 days, and refreezing was avoided. The copy number of each ICO was calculated from the concentration data provided by the manufacturer. The optimum concentration of ICO was determined separately for each assay in titration studies (data not shown). The amount of ICO in the reaction was kept as low as possible to minimize competition in low-titer samples and to detect weak inhibition or low efficiency of the nucleic acid extraction by failure of ICO amplification. For the HBV, HSV, and VZV assays, 6000–62 500 ICO copies/mL, corresponding to 60–625 copies/reaction, were added directly to the sample before DNA extraction. For the MTBC assay 1500 ICO copies/reaction were included in the reaction mixture.
dna extraction
DNA of M. tuberculosis was isolated with the AMPLICOR respiratory specimen preparation reagents according to the manufacturer’s instructions (Roche Diagnostics). Briefly, after the sample was decontaminated with NaOH–N-acetylcysteine, 100 µL of the sample was mixed with 1 mL of washing buffer RW. After centrifugation the pellet was resuspended in 100 µL of alkaline lysis buffer RL and incubated at 60 °C for 45 min. Buffer RN (100 µL) was added to the sample to adjust it to neutral pH.
VZV, HSV1, and HSV2 DNAs were extracted from lyophilized samples purchased for proficiency testing (INSTAND). Each lyophilized sample was first dissolved in 1 mL of water. The DNA was then extracted from 200 µL of the reconstituted sample with the QIAamp DNA Blood Mini Kit (Qiagen), according to the manufacturer’s instructions. The same reagents were used to isolate HBV DNA from 200 µL of plasma.
protocol for pcr and melting analysis
Rapid-cycling PCR and melting analysis were performed in a LightCycler. PCR was done in glass capillaries (Roche Molecular Biochemicals) that contained either 20 µL (MTBC assay) or 10 µL (all other assays) of reaction mixture. Each 10 µL of reaction mixture contained: 5 pmol of each primer, 2 pmol each of the anchor and detection probes, 2 µL of LightCycler-FastStart DNA Master Hybridization ProbesPLUS (which includes reaction buffer, nucleotides, and Taq polymerase; Roche Molecular Biochemicals), and 2.5 µL of DNA.
PCR amplification for the HBV assay was performed as described previously (8). For fluorescence acquisition during amplification, the temperature of the reaction was increased to 60 °C. A melting point analysis (see below) was performed after amplification.
The thermocycling conditions for MTBC, HSV, and VZV amplification were as follows: 95 °C for 10 min for initial denaturation and activation of Taq polymerase, followed by 50 thermal cycles of 95 °C for 10 s, 55 °C for 10 s, 60 °C for 5 s, and 72 °C for 10 s, with a ramping rate of 20 °C/s. Fluorescence was measured during each 60 °C stage.
After amplification, the instrument performed a melting analysis by heating the capillary at 95 °C for 10 s, incubating it at 45 °C for 1 min, and then slowly (0.2 °C/s) heating it to 80 °C. Fluorescence was monitored continuously during the melting experiment. To convert melting curves to melting peaks, the LightCycler software (Ver. 3.5; Roche Molecular Biochemicals) calculated the negative derivative of each measured fluorescence with respect to the temperature (–dF/dT), and then plotted –dF/dT against temperature for the entire melting experiment.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, sample 2). However, the subsequent melting point analysis starts at a temperature lower than the melting temperature of the detection probe–ICO hybrids. Therefore, during melting point analysis, a control-specific fluorescence signal (melting peak) is obtained as these hybrids melt out. For example, Fig. 1C
shows that negative control sample 2 has a melting peak at 50 °C. In contrast, positive samples (e.g., sample 1 in Fig. 1
) showed both an increase in fluorescence during real-time amplification (Fig. 1B
) and a target-specific melting peak (65 °C in this case) during melting point analysis (Fig. 1C
).
|
Note that, depending on the concentration of the target DNA, a second, ICO-specific peak (melting peak at 50 °C in sample 1, Fig. 1C
) could also be visible in positive samples. However, because only a low amount of ICO is used in an assay, the ICO is not amplified in most positive samples because it competes (unsuccessfully) with the target DNA.
The LightCycler real-time PCR assays to which we have successfully added ICOs are listed in Table 1
. Table 1
also lists the melting temperatures of the hybrids formed between the detection probes and the templates and ICOs. In all of these examples the difference between the temperatures of these hybrids was >5 °C. This temperature difference prevents the control from producing detectable fluorescence at the measuring step (during amplification) and allows the control and the target to be clearly distinguished during the melting analysis.
ico-controlled real-time pcr assay for the detection of mtbc
To detect the influence of the ICO on the sensitivity of the assay, we assayed a dilution series of MTB DNA with and without internal control (Fig. 2
). As shown in Fig. 2A
, there were no detectable differences in sensitivity between the sample with 5000 genome copies/reaction and the sample with the lowest concentration of the dilution series (25 copies/reaction). In the melting analysis (Fig. 2B
), both the negative sample (curve 5) and the positive sample with low MTB copy number (curve 4) had an ICO peak at 49 °C. In the positive sample with the higher number of target copies (curve 2), the ICO was not amplified because of competition between target and control DNA.
|
Because both hybridization probes exhibited similar melting temperatures, no distinct MTBC melting peak could be detected (Fig. 2B
). However, the assays usually showed a broad peak between 59 and 65 °C, and the shape of the peak changed with the amount of template DNA. In all cases, the MTBC melting peak could be clearly differentiated from the distinct ICO peak at 49 °C.
The positive MTBC PCR results obtained with the ICO-controlled LightCycler assay for five clinical samples that had previously been tested by microscopy, culture, and two different commercial assays (Roche Amplicor and Becton Dickinson ProbeTec) are shown in Fig. 3
. Two of the samples (curves 3 and 5) had tested negative in microscopy, indicating that they had a low titer of mycobacterial cells. The ICO-specific peak at 49 °C is present in the negative sample (Fig. 3B
, line 6) and in two of the positive samples (Fig. 3B
, lines 3 and 5). Detailed features of the new LightCycler MTBC assay will be described elsewhere (Burggraf et al., manuscript in preparation).
|
application of icoS for other diagnostic pcrS
To show the feasibility of the ICO technique, we added the appropriate ICOs (Table 1
) to existing real-time PCR assays for three different pathogens: HBV (8), HSV(11), and VZV (11). For the HBV assay (Fig. 4
), the test samples were a dilution series derived from a highly positive sample. The HBV ICO (62 500 copies/mL) was added to each dilution before DNA extraction. A fluorescence signal was detected from samples with starting concentrations as low as 1000 copies/mL (corresponding to 10 copies of HBV DNA per PCR reaction; Fig. 4A
, curve 3). This is identical to the sensitivity obtained with the assay that did not contain ICO (data not shown). No fluorescence signal was detected from the reaction containing 1 copy of HBV (corresponding to 100 copies/mL of plasma; Fig. 4A
, curve 4).
|
To simulate inhibition, we added 100 g/L ethanol to a reaction containing 10 copies of HBV DNA. No fluorescence signal was obtained from this reaction (Fig. 4A
, curve 5) although a signal was clearly visible in the reaction without inhibitor (curve 3). Melting point analysis (Fig. 4B
) revealed a HBV-specific melting peak at 65 °C for all positive samples. An ICO-specific peak at 50 °C was obtained from the two samples with lower HBV copy numbers (Fig. 4B
, curves 2 and 3) as well as from two negative samples [100 copies/mL (curve 4) and negative control (curve 6)]. Because of competition, no ICO peak was seen in the high-titer HBV sample (curve 1). No amplification of the ICO occurred in the reactions containing the inhibitor (curves 5 and 7), clearly showing that the result obtained from sample 5 (Fig. 4A
) was a false negative.
Similar experiments with assays targeted to HSV and VZV were also successful. For these, the assays originally described by Stöcher et al. (11) were supplemented with ICOs (Table 1
). The melting point analysis for HSV2 (curve 1), HSV1 (curve 2), and a negative sample are shown in Fig. 5
.
|
The melting temperature of detection probe–HSV2 hybrids was 72 °C. Because of two mismatches in the binding region for the detection probe, the melting temperature for probe–HSV1 hybrids was lower (60 °C). A strong ICO peak at 52 °C was seen in the negative sample.
The melting temperatures measured with the VZV assay were 60 and 52 °C for target and ICO, respectively (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In contrast to the hypothesis that PCR amplification efficiency mainly depends on the size and GC content of the template DNA (13), experiments have indicated that primer sequences are the major factor in amplification efficiency (7)(14). In this study we demonstrated that single-stranded oligonucleotides, consisting of little more than primer- and probe-binding regions, can serve very well as internal controls. The amplified ICOs are detected by the same probe that detects the amplified target. We successfully applied this technique to a newly designed LightCycler assay for rapid detection of MTBC and to three virus-specific PCR assays published by others (8)(11). We have routinely used ICOs for several months without any major problems. In only few cases have we noticed a failure or very weak ICO amplification, which is probably attributable to deterioration of the ICO after several cycles of refreezing.
A limitation of the described technique is that ICOs cannot be used as internal standards in quantitative PCR assays, which are performed with internal controls detected by a separate probe. However, if this is required, external calibrators can be used to obtain exact quantification (15)(16). Another limitation may be the inability to detect partial PCR inhibition, which could influence quantitative PCR results. In an assay with a separate control-specific probe, partial inhibition can be detected by a late increase of the fluorescence signal derived from the control. This is not possible with the ICO method because no real-time amplification signal is produced. However, this disadvantage can be minimized by careful titration of the ICO. The amount of ICO in the reaction mixture should be as low as possible to detect even weak inhibition.
These limitations are counterbalanced by several advantages of our single-probe approach. The ICO method is cost-effective because a second, control-specific probe is not necessary. One technical advantage is that separate probes do not need to be added to the reaction mixture, which decreases the chance for oligonucleotide dimerization and possibly increases the sensitivity of the assay. Furthermore, an assay with a separate control-specific probe requires an external positive control to check the function of the target-specific probe. Moreover, in real-time amplification instruments, only a limited number of analysis channels are available for the detection of different fluorescent dyes, e.g., the original LightCycler instrument had only two available channels. Therefore, if a laboratory wanted to develop a multiplex PCR application to detect two or more different pathogens in a single reaction vessel, it would be a disadvantage to require that one detection channel be reserved for the internal control.
In multiplex applications, it also would seem possible to add several different ICOs, each specific for one of the targets of the assay, thereby providing controls for all components of the reaction mixture. Without ICOs, this could be achieved only by adding separate external positive controls for each of the target nucleic acids to the experiment.
Although it is well recognized that a performance control is a prerequisite for reliable PCR, to date most published PCR assays do not contain an internal amplification control (12). Many diagnostic laboratories lack the time, equipment, and facilities to perform the cloning experiments necessary to generate plasmid controls. They could, however, design suitable ICOs that could be synthesized by someone else. Commercial providers usually can handle high-quality synthesis of oligonucleotides longer than 100 bases. Prices are generally reasonable, and even if a laboratory ordered the minimum amount that the providers would be willing to synthesize, the yield (in terms of assays) would be substantial. For example, only 5 nmol of purified ICO would be enough to perform 
3 x 1012 assays if 1000 copies of the oligonucleotide were added as internal control to each assayed sample before DNA extraction.
In summary, in this study we successfully applied the new ICO approach to the detection of different pathogens. Our technique is suitable for all existing real-time amplification protocols that allow a melting point analysis after amplification. Introduction of ICOs could further increase the reliability of nucleic acid amplification assays.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. Stormer, K. Kleesiek, and J. Dreier High-Volume Extraction of Nucleic Acids by Magnetic Bead Technology for Ultrasensitive Detection of Bacteria in Blood Components Clin. Chem., January 1, 2007; 53(1): 104 - 110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Espy, J. R. Uhl, L. M. Sloan, S. P. Buckwalter, M. F. Jones, E. A. Vetter, J. D. C. Yao, N. L. Wengenack, J. E. Rosenblatt, F. R. Cockerill III, et al. Real-Time PCR in Clinical Microbiology: Applications for Routine Laboratory Testing Clin. Microbiol. Rev., January 1, 2006; 19(1): 165 - 256. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Burggraf and B. Olgemoller Straightforward Procedure for Internal Control of Real-Time Reverse Transcription Amplification Assays Clin. Chem., August 1, 2005; 51(8): 1508 - 1510. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Burggraf, U. Reischl, N. Malik, M. Bollwein, L. Naumann, and B. Olgemoller Comparison of an Internally Controlled, Large-Volume LightCycler Assay for Detection of Mycobacterium tuberculosis in Clinical Samples with the COBAS AMPLICOR Assay J. Clin. Microbiol., April 1, 2005; 43(4): 1564 - 1569. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. S. Nolte Novel Internal Controls For Real-Time PCR Assays Clin. Chem., May 1, 2004; 50(5): 801 - 802. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), 5000 copies of MTB with 1500 copies of the ICO; curve 3 (
), 25 copies of MTB/reaction without ICO; curve 4 (
), 25 copies of MTB/reaction with 1500 copies of the ICO; curve 5, MTB-negative sample with 1500 copies of the ICO (sterile water copurified with the samples). (A), fluorescence signal obtained during amplification; (B), melting point analysis.
