Calibrated real-time PCR for evaluation of parvovirus B19 viral load.
Fluorescence-based real-time PCR assays can combine specific detection and quantitative evaluation of the viral load. For quantitative evaluation of the viral load, realtime PCR assays must be carefully designed to give reliable results. In particular, the mode of quantification, whether absolute or relative, and an appropriate calibration method need to be firmly established (4, 5). Absolute quantification can be obtained by referring to a calibration curve. To take into account the variability attributable to cumulating effects both in sample processing and in the analytical phases, the calculated amount of target nucleic acids can be normalized to the calculated amount of a reference target that is likely to be present in constant amounts in all samples analyzed. Alternatively, relative quantification can be obtained from the ratio of target present in experimental samples to target in a calibrator sample. To compensate for sample variability, ratios can be further normalized with respect to a parallel, relative quantification of a reference target (6).
Different real-time PCR assays for the detection of B19 virus in clinical samples have been developed and described. These include use of the LightCycler system and detection by use of Sybr Green (7) or fluorescence resonance energy transfer probes (8), as well as by use of the ABI Prism system and detection by TagMan probes (9,10). In our work, we developed a calibrated real-time PCR assay that uses the LightCycler system for the detection and quantitative evaluation of B19 virus nucleic acids. B19 virus target in experimental samples was quantified relative to a calibrator sample containing a known amount of viral target. An exogenous nucleic acid was designed as an analytical standard for monitoring sample processing and amplification and as a reference target to normalize quantification.
For the proposed real-time PCR assay, we constructed two plasmids. Plasmid pHRO contains a 4900-bp viral segment, corresponding to nucleotides 346-5245 of the B19 virus genome (National Center for Biotechnology Information Genome Database NC 000883), inserted in plasmid vector pGEM3Z (Promega). Plasmid pEC01 contains a 192-bp synthetic segment, designed as a unique sequence unrelated to known sequences present in the National Center for Biotechnology Information nucleotide database (5'-ttttcttgttatttctcctgctacttgtcgtggtagttatcatgataacctcgtcatcttcccc gccacctcccgcggcagctgccacgacaactaggtgat-gttgctggcgacgtcgccggggaggtggcgggagaagt agcaaatattactagtaacatcaccagcaagatgacgaggaaaataacaaga-3'), inserted in plasmid vector pTRIl9amp (Ambion). Both viral (HRO) and reference (EC01) target DNA were produced as linear amplification products from template plasmid DNA and used as genome equivalents (geq) standards in the following amplification reactions. Specific primer pairs were designed for the real-time PCR assay that were optimal in terms of specificity, sensitivity, and efficiency of amplification (see Table 1 in the Data Supplement that accompanies the online version of this Technical Brief at http: / / www.clinchem.org/content/vol50/issue4/).
Real-time PCR was performed with the LightCycler system and Sybr Green detection of amplification products (see Table 2 in the online Data Supplement). For evaluation of the real-time PCR assay, fluorescence emission was analyzed for target quantification and melting curve profiles. Analysis of the amplicon accumulation curves by use of the fit-point method yielded, for each reaction, crossing point (Cp) numbers that were related to the log of target copy number with the function present into the software. Analysis of the melting curve profiles confirmed specific accumulation of the amplification products, with a single melting peak for the viral target [melting temperature (T = 82.6-82.8 [degrees]C] and a double melting peak for the reference target (Tml = 79.6-79.8 [degrees]C; T = 85.8-86.2 [degrees]C). Specific amplification of the viral and reference targets was further confirmed by sequencing of amplification products.
The characteristics of the assay have been evaluated according to guidelines for the validation of analytical procedures (11). Serial dilutions of HRO or EC01 target sequences, from [10.sup.2] to [10.sup.8] geq/reaction, were amplified in triplicate series repeated three times in different experimental sessions. Calibration curves were constructed by plotting Cp values as a function of log of target copy number. For both the viral and reference targets, the calibration curves obtained were linear in the range [10.sup.2] - [10.sup.8] geq. The detection limit, determined by endpoint dilution analysis, was [10.sup.1] geq. Considering the linear range interval, these curves showed high correlation, accuracy, and precision as reported (see Table 3 in the online Data Supplement). The efficiencies of the two amplification reactions were calculated from the slopes of linear regression curves following the relationship: E = [10-.sup.(-1/slope)] and were 1.872 for HRO and 1.638 for EC01.
In the following experiments, the viral target in experimental samples was quantified relative to a calibrator sample and normalized to the reference target. Quantification was obtained by applying the relationship (6): [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] target ref target and Eref are the efficiencies of the amplification reactions for viral and reference targets as determined previously, and OCptarget and OCpref are the deviations in respective Cp values, obtained in parallel amplification reactions, between a control sample, used as calibrator and the sample analyzed. Absolute quantification of the viral target was then obtained by multiplying for the known amount of viral target present in the calibrator.
The characteristics of this model for quantification were first evaluated by amplifying serial dilutions of HRO target sequences, from [10.sup.1] to [10.sup.8] geq/reaction, in the presence of [10.sup.3] geq of EC01 target sequences in triplicate series repeated three times in different experimental sessions. Samples were separately amplified with the specific primer pairs, and Cp values were obtained for both viral and reference targets. The mean standard deviation (and its SD) for the grouped Cp values for the viral target in the range [10.sup.2]-[10.sup.8] geq was 0.47 (0.19); the mean (SD) standard deviation for the Cp values for the reference target at [10.sup.3] geq was 0.20 (0.09). We therefore applied the aboveindicated relationship, considering as calibrator each sample in turn, and performed regression analysis on logtransformed data to correlate the obtained experimental values with the expected values. Quantification of the viral target normalized with respect to the reference target led to experimental values in good agreement with expected values, independent of the sample used as calibrator (Table 1 and Fig. 1).
For the analysis and quantitative evaluation of B19 viral load in serum and plasma specimens, the International Standard for B19 DNA (NIBSC 99/800; [10.sup.9] IU/L) was used as the standard sample (12). Serial dilutions of the International Standard, containing [10.sup.5] to [10.sup.8] IU/L, were processed in the presence of [10.sup.7] geq/L EC01 DNA by a lysis and precipitation procedure (see the Note in the online Data Supplement), and duplicate samples were amplified separately with the specific primer pairs for both viral and reference targets. When we considered as calibrator any dilution of the International Standard processed together with EC01 reference target DNA, relative quantification yielded experimental values, directly expressed as IU/L, that showed linearity with expected values in the detection interval from [10.sup.6] to [10.sup.8] IU/L. The mean (SD) values for the parameters obtained by regression analysis on log-transformed data were as follows: slope = 1.001 (0.038); [r.sup.2] = 0.863; SD of the residuals = 0.176.
The experimental procedure and the analytical model were therefore applied in two different situations in which determination of B19 viral load is especially relevant: clinical evaluation in the course of prolonged viremia, and prevention of transmission of the virus.
Serum samples were obtained consecutively during a 12- to 16-month period from three different individuals with documented persistent B19 virus in the peripheral blood associated with clinical symptoms typical of B19 virus infection, such as recurrent cutaneous rash and arthropathies. As control, the International Standard at the dilution of [10.sup.8] IU/L was used. All sera and controls were processed in the presence of [10.sup.7] geq/L EC01 DNA. All samples were then analyzed by real-time PCR for quantification of the viral target and normalized to the reference target. Viral load values obtained by quantitative real-time PCR indicated persistence of the virus in the peripheral blood of these individuals at significant concentrations for extended periods of time (see Fig. 1 in the online Data Supplement).
We also analyzed 16 plasma pool samples, each derived from 960 single donations and representative of manufacturing plasma pools for the production of blood derivatives (obtained from Kedrion Biopharmaceuticals, Bolognana, Lucca, Italy), using as control the International Standard at [10.sup.8] IU/L. All samples and controls were processed and analyzed as described, and viral loads were obtained (see Fig. 2 in the online Data Supplement). In plasma pools for the production of blood derivatives, quantitative evaluation of the viral load is required as a manufacturing standard to reduce the risk of transmission of virus, and a threshold considered acceptable for contamination has been set at [10.sup.7] geq/L (3). B19 virus can be present in the blood at very high concentrations, and because of the frequency of viremic individuals in a donor population, contaminated blood might be inadvertently included in the manufacturing of plasma pools (13), leading to measurable contamination of the final blood products (14). Of the 16 plasma pools, only 2 did not have detectable B19 virus, whereas 4 had viral loads higher than the threshold limit of [10.sup.7] IU/L.
In this study, we obtained quantitative results by use of a mathematical model that quantified the viral target present in experimental samples with respect to a calibrator sample. A synthetic nucleic acid added in defined amounts to the sample at the start of the analytical process was appropriate both as an analytical standard and as a reference target for normalization of relative amounts. Because in our protocol the calibrator sample contained known amounts of viral and reference target sequences, the relative value could be converted to an absolute value for target viral load in the samples.
[FIGURE 1 OMITTED]
This work was supported by funds from Ministero dell'Istruzione, dell'Universita e della Ricerca and from the University of Bologna. The LightCycler instrument was made available by Centro Interdipartimentale Ricerche Biotecnologiche, University of Bologna.
(1.) Zerbini M, Musiani M. Human parvoviruses. In: Murray PR, Baron EJ, Jorgensen JH, Pfaller MA, Yolken RH, eds. Manual of clinical microbiology, 8th ed. Washington: ASM Press, 2003:1534-43.
(2.) Gallinella G, Zuffi E, Gentilomi G, Manaresi E, Venturoli S, Bonvicini F, et al. Relevance of B19 markers in serum samples for a diagnosis of parvovirus B19-correlated diseases. J Med Virol 2003;71:135-9.
(3.) Tabor E, Epstein JS. NAT screening of blood and plasma donations: evolution of technology and regulatory policy. Transfusion 2002;42: 1230-7.
(4.) Rasmussen R. Quantification on the LightCycler. In: Meuer S, Wittwer C, Nakagawara K, eds. Rapid cycle real-time PCR: methods and application. Berlin: Springer, 2001:21-34.
(5.) Mackay IM, Arden KE, Nitsche A. Real-time PCR in virology. Nucleic Acids Res 2002;30:1292-305.
(6.) Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001;29:2002-7.
(7.) Manaresi E, Gallinella G, Zuffi E, Bonvicini F, Zerbini M, Musiani M. Diagnosis and quantitative evaluation of parvovirus B19 infections by real-time PCR in the clinical laboratory. J Med Virol 2002;67:275-81.
(8.) Harder TC, Hufnagel M, Zahn K, Beutel K, Schmitt H-J, Ullmann U, et al. New LightCycler PCR for rapid and sensitive quantification of parvovirus B19 DNA guides therapeutic decision-making in relapsing infections. J Clin Microbiol 2001;39:4413-9.
(9.) Aberham C, Pendl C, Gross P, Zerlauth G, Gessner M. A quantitative, internally controlled real-time PCR Assay for the detection of parvovirus B19 DNA. J Virol Methods 2001;92:183-91.
(10.) Gruber F, Falkner FG, Dorner F, Hammerle T. Quantitation of viral DNA by real-time PCR applying duplex amplification, internal standardization, and two-color fluorescence detection. Appl Environ Microbiol 2001;67:2837-9.
(11.) International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). Guidelines for validation of analytical procedures (Q2A) and methodology (Q2B). http://www.ich.org. (Q2A, October 1995; Q2B, November 1996).
(12.) Saldanha J, Lelie N, Yu MW, Heath A. Establishment of the first World Health Organization International Standard for human parvovirus B19 DNA nucleic acid amplification techniques. Vox Sang 2002;82:24-31.
(13.) Gallinella G, Moretti E, Nardi G, Zuffi E, Bonvicini F, Bucci E, et al. Analysis of B19 virus contamination in plasma pools for manufacturing, by using a competitive polymerase chain reaction assay. Vox Sang 2002;83:324-31.
(14.) Siegl S, Cassinotti P. Presence and significance of parvovirus B19 in blood and blood products. Biologicals 1998;26:89-94.
Giorgio Gallinella, Francesca Bonvicini, Claudia Filippone, Stefania Delbarba, Elisabetta Manaresi, Marialuisa Zerbini, and Monica Musiani
(Department of Clinical and Experimental Medicine, Division of Microbiology, University of Bologna, Bologna, Italy; * address correspondence to this author at: Department of Clinical and Experimental Medicine, Division of Microbiology, University of Bologna, Via Massarenti 9, I-40138 Bologna, Italy; fax 39-051-341632, e-mailggallinemed.unibo.it)
Table 1. Linear regression analysis of quantitative values obtained with respect to diverse calibrator samples, each containing the indicated amount of viral target in the presence of [10.sup.3] geq of EC01 reference target. (1) Calibrator, geq Slope SD of slope [10.sup.8] 1.008 (0.013) 0.006 (0.002) [10.sup.7] 0.999 (0.007) 0.006 (0.002) [10.sup.6] 0.989 (0.001) 0.006 (0.001) [10.sup.5] 0.991 (0.004) 0.006 (0.001) [10.sup.4] 1.009 (0.013) 0.006 (0.002) [10.sup.3] 1.019 (0.009) 0.006 (0.003) [10.sup.2] 0.985 (0.025) 0.006 (0.002) Mean 1.000 (0.010) 0.006 (0.002) Calibrator, SD of the geq [R.sup.2] residuals [10.sup.8] 0.996 (0.003) 0.128 (0.042) [10.sup.7] 0.997 (0.002) 0.116 (0.040) [10.sup.6] 0.997 (0.001) 0.120 (0.026) [10.sup.5] 0.996 (0.002) 0.123 (0.026) [10.sup.4] 0.997 (0.002) 0.112 (0.041) [10.sup.3] 0.996 (0.003) 0.124 (0.051) [10.sup.2] 0.997 (0.002) 0.113 (0.040) Mean 0.996 (0.002) 0.119 (0.038) (a) Mean (SD) values were determined from triplicate series, repeated three times in different experimental sessions.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Technical Briefs|
|Author:||Gallinella, Giorgio; Bonvicini, Francesca; Filippone, Claudia; Delbarba, Stefania; Manaresi, Elisabe|
|Date:||Apr 1, 2004|
|Previous Article:||Diminished urinary free cortisol excretion in patients with moderate and severe renal impairment.|
|Next Article:||Reevaluation of the power of error detection of Westgard multirules.|