Previous Article | Next Article ![]()
Journal of Clinical Microbiology, April 2005, p. 1738-1744, Vol. 43, No. 4
0095-1137/05/$08.00+0 doi:10.1128/JCM.43.4.1738-1744.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Department of Medical Microbiology,1 Department of Pediatrics, Leiden University Medical Center, Leiden, The Netherlands,2 Department of Pediatrics, Motol University Hospital, Prague, Czech Republic,3 Centro de Educacion Medica e Investigaciones Clinicas, Buenos Aires, Argentina4
Received 21 July 2004/ Returned for modification 25 August 2004/ Accepted 17 December 2004
|
|
|---|
|
|
|---|
Several quantitative real-time PCR assays for monitoring herpesvirus DNA load have been used in the follow-up for patients posttransplantation, resulting in successful implementation of preemptive therapy for CMV and EBV infections (2, 26, 32). Our preliminary studies have shown that adenovirus DNA in serum from immunocompromised patients may be predictive for the outcome of disease (7). Application of a semiquantitative PCR assay for the detection of adenovirus DNA in the plasma of transplant patients suggested an association of the viral load and fatal outcome of the disease (29).
Recently, real-time PCR applications for the detection and quantification of adenovirus have been described for species C viruses in transplant patients (10, 34) and adenovirus type 4 for respiratory diagnosis (16) and also for broadly reactive adenovirus multiplex assays on a variety of clinical samples (12, 14, 23, 24).
In this study, an internally controlled, multiplex real-time PCR that enabled efficient quantification of species A, B, and C adenoviruses was developed. An internal control is coamplified to monitor the integrity of the DNA extracted from clinical samples as well as the presence of PCR inhibitors. The heterogeneity of the adenovirus family complicates the design of an adenovirus real-time PCR. Mismatches in primer and probe sequences can significantly affect the kinetics of a real-time PCR and thus quantification. Our approach has focused on a PCR that amplified DNA of all relevant adenovirus serotypes by using primers selected from the 3' end of the hexon gene (5, 6) and the detection of the amplified products by using molecular beacons as probes (35). By using molecular beacons carrying different fluorophores, species of adenoviruses could be differentiated in the assay (37). Laboratories lacking multicolor real-time PCR detection can apply the assay with the same primers but using a minor groove binder (MGB) probe (19). This new assay has been applied retrospectively on a series of pediatric allogeneic stem cell transplant (SCT) patients previously tested by semiquantitative analysis (29). Since 2002, this assay has also been prospectively applied to monitor adenovirus DNA in allogeneic SCT patients (20) and to evaluate the efficacy of ribavirin treatment of this condition (21). This assay contributes to better management of adenovirus infections in transplant patients.
|
|
|---|
An adenovirus type 5 stock of 2.34 x 1011 particles/ml as counted by electron microscopy was purchased from Advanced Biotechnologies Inc, Columbia, Md.
Two DNA plasmids containing an insert of the 3' end region of the hexon gene of adenoviruses type 5 (IQL 103) and type 35 (IQL 104) were constructed (IQ Products, Groningen, The Netherlands). The concentration was 108 copies of DNA/ml.
Primers and probes. The primers previously published by Echavarria et al. (5) were adjusted to better suit real-time PCR kinetics. These primers are selected from the 3' end of the adenovirus hexon gene and resulted in a PCR product of 137 bp. Species-specific molecular beacons carrying different fluorophores were designed for detection and quantification of the different species of adenoviruses. Primers and probes are shown in Table 1.
|
View this table: [in a new window] |
TABLE 1. Primers and probes for real-time PCR detection and amplification of adenoviruses
|
In addition, a minor groove binder (MGB) probe was designed. This probe can be used instead of the molecular beacons for general detection and quantification of the same adenovirus PCR products, without differentiation of the species. The MGB probe was synthesized by Applied Biosystems Inc, Nieuwerkerk aan den IJssel, The Netherlands.
Patients. Allogeneic SCT recipients at the Department of Pediatrics in the Leiden University Medical Center were included in the study, and a unique patient number (UPN) identified the patients. The original diseases necessitating the transplantation included acute myeloblastic leukemia, acute lymphoblastic leukemia, myelodysplastic syndrome, severe combined immunodeficiency disorder, severe aplastic anemia, Fanconi aplastic anemia, and chronic myelomonocytic leukemia. Occurrence of adenovirus infections was not related to the original disease.
In a retrospective study of 328 allogeneic SCT patients, 36 patients had a positive adenovirus culture from stool, urine, or a throat swab and were selected for further analysis as described previously (29). These patients were divided into three groups based on their clinical presentation. Group I had positive adenovirus cultures but no clinical symptoms (n = 17). Group II had adenovirus-related diseases but survived (n = 12). Ten of these patients had enteritis, one patient had hemorrhagic cystitis, and one patient had enteritis and hemorrhagic cystitis. Group III comprised patients with a fatal outcome that clinically was attributed to a disseminated adenovirus infection (n = 7). All patients but one had enteritis in combination with hepatitis or hemorrhagic cystitis. The remaining patient had pneumonia.
Subsequently, in 2001 and 2002, 48 allogeneic transplant recipients were monitored prospectively for 6 months after SCT.
Extraction of DNA. DNA was isolated from sera, plasma samples, and controls by using spin columns of the QiaAmp blood kit (QIAGEN GmbH, Hilden, Germany) according to the manufacturer's instructions.
For the prospective study, DNA was isolated from clinical samples, usually plasma, by the Total Nucleic Acid isolation kit and the automated MagNApure LC DNA extraction system (Roche Diagnostics, Mannheim, Germany). No significant differences were observed between the results obtained by column isolation and those obtained by automated extraction in the MagNApure system (data not shown).
Real-time PCR assay. Real-time PCR was performed using the iCycler IQ multicolor real-time PCR detection apparatus (Bio-Rad, Veenendaal, The Netherlands), which enables simultaneous detection of four different fluorophores. Binding of the specific probes to the PCR products generates fluorescence, which is detected by the iCycler IQ. The amplification cycle at which this fluorescence crosses the threshold line is called the threshold cycle or CT value.
The PCR is performed in a 50-µl reaction volume containing 2x HotStar Taq Master mix (QIAGEN GmbH) as a multiplex assay with three primers (0.3 µM concentrations of each primer [ADVs, ADV31s, and ADVas]), two probes (a 3 µM concentration of 6-carboxyfluorescein [FAM]-labeled ADV-MB-AC for species A and C virus detection and a 3 µM concentration of hexachlorofluorescein [HEX]-labeled ADV-MB-B for species B virus detection), and a final concentration of 4 mM MgCl2. Addition of 10 µl of extracted DNA completed the reaction mixture. The amplification protocol consisted of an enzyme activation step of 95°C for 15 min followed by 50 cycles of amplification (95°C for 30 s, 55°C for 30 s, and 72°C for 30 s). A fixed amount of PhHV was added to the lysis buffer prior to extraction and was coamplified in the reaction to check DNA isolation and inhibition as described previously (33).
Quantification of the adenovirus load. The extraction of adenovirus control of 108 VP/ml was identical to that of the patient samples, and dilution series of 102 to 107 VP/ml were applied to real-time PCR. The obtained CT values were used to establish a standard curve, and subsequently the viral load of the clinical samples was determined by entering the CT value of the sample in the standard curve. The software of the iCycler IQ multicolor real-time PCR detection system performed the calculations.
Sequence analysis. Two additional primers (sense, GAAACTTCCAGCCCATGAG; antisense, TTATGTGGTAGCGTTRC) were used for nucleotide sequence analysis of an approximately 450-bp fragment of the 3' end hexon gene. In this way, the primer and probe regions of adenovirus serotypes were analyzed to define the appropriate primers, and in addition, this method was used for genotyping the viruses. The PCR products were subjected to sequence analysis with the Big Dye-Terminator Cycle Sequencing ready reactions kit containing Ampli-Taq DNA polymerase FS (Applied Biosystems, Inc., Foster City, Calif.). Samples were analyzed on an ABI model 310 DNA sequencer (Applied Biosystems, Inc.), and genotyping was performed by BLAST analysis on the NCBI website (www.ncbi.nlm.nih.gov/BLAST).
|
|
|---|
|
View this table: [in a new window] |
TABLE 2. Nucleotide sequences of the primer and probe regions at the 3' end of the hexon gene of the subgroup A, B, and C serotypes of adenovirus
|
|
View this table: [in a new window] |
TABLE 3. Inefficient amplification of HAdV 31 shown by heterogeneity of viral DNA loads of experimental dilution series of adenovirus and heterologous serotypes as standarda
|
An internal control reaction was coamplified in the assay to check the procedure for efficient DNA isolation and for inhibition. A fixed amount of PhHV was added to the clinical samples prior to DNA isolation and coamplified in the real-time PCR using a Cy5-labeled probe for detection. Only if the CT value for this spike was within two cycles of the average of uninhibited samples was the result of the specimen considered a valid result.
Quantification and species. The heterogeneity in the adenovirus family complicates standardized quantification, as shown in Table 3, due to primer mismatches. When analyzing dilution series, differences were observed in the CT values of similar concentrations of HAdV 5 (species C) and HAdV 35 (species B) HPLC-titrated virus standards. This result most likely relates to the fact that in HPLC titration, the quantity is based on particles rather than DNA copies. For that reason, plasmid DNA standards were prepared and it was shown that the amount of DNA of this plasmid stock of HAdV 5 was comparable to the values for the HAdV 5 virus standards. This was further confirmed by obtaining a commercial batch of HAdV 5, which consisted of purified virus, that was quantified by electron microscopy (Advanced Biotechnologies). Therefore, the subgroup C standard was used routinely in the assay and allowed for accurate quantification of all serotypes of the A and C subgroup viruses, which are predominantly associated with adenovirus complications.
When a subgroup B virus was detected in the multiplex PCR, the sample was retested with the specific, less well standardized standard curve of that subgroup.
The MGB probe can be used for quantification of adenoviruses as well. No significant difference in viral loads was obtained when using this probe instead of the molecular beacons, provided that the probe was also used for the standard curve amplification.
Sensitivity and specificity. The sensitivity of the assay was determined by using the titrated HAdV 5 and HAdV 35 stocks and the two plasmids containing the 3' end of the hexon genes of these subtypes. Dilution series of these standards were subjected to real-time PCR amplification. The CT value was plotted against the dilution to generate a standard line that can be used for the quantification of unknown samples by CT value. The sensitivity of the assay was shown to be between 50 and 250 copies/ml, which correlates to 1 to 5 copies in the reaction tube (data not shown).
The amplification was also shown to be specific for adenoviruses. No amplification signal was obtained when nucleic acids from CMV, EBV, human herpesvirus 6, parvovirus, hepatitis B virus, hepatitis C virus, human immunodeficiency virus, and BK virus were used as targets.
Retrospective analysis of patients. Pediatric stem cell transplant recipients from a retrospectively analyzed cohort with documented adenovirus infections had been divided into groups I, II, and III based on increasing severity of infection by clinical interpretation (29). DNA was extracted from serum samples and subjected to real-time PCR. Results of the DNA load in the samples from the adenovirus-positive patients are shown in Table 4. In four out of six patients with fatal disseminated infection (group III), a steep increase of the adenovirus DNA load in serum could be observed in the month preceding their death, with a maximum of 1012 copies/ml in UPN 321. Of these four patients, two suffered from combined enteritis and hepatitis, one had enteritis and hemorrhagic cystitis, and one had encephalitis. Unfortunately, no serum samples were available from UPN 369 (pneumonia) and UPN 376 (enteritis and hepatitis) from the weeks preceding their deaths.
|
View this table: [in a new window] |
TABLE 4. Retrospective monitoring of adenovirus DNA in serum samples of SCT recipients pre and posttransplantation
|
In 2 out of 12 nonfatal cases (group II) and even in 5 out of 17 asymptomatic patients (group I), adenovirus DNA could be occasionally detected in serum as well, sometimes even before transplantation. Usually the load did not exceed 105 copies/ml, but in the asymptomatic patients UPN 168 and UPN 286, DNA loads of 106 to 107 copies/ml were observed.
Apart from one case of HAdV 41 viremia, only adenovirus species A (comprising serotypes 12, 18, and 31) and species C (serotypes 1, 2, 5, and 6) have been detected in this patient group, with the latter being more prevalent (Table 4).
Prospective study of adenovirus DNA loads. Using the real-time PCR protocol, 48 pediatric patients receiving transplants from January 2001 until December 2002 were monitored prospectively. The available plasma samples were checked for adenovirus DNA for at least 6 months after transplantation. In less than 0.5% (n = 4) of over 800 specimens, inhibition of the PCR was observed, as shown by reduced PhHV internal control amplification.
In 12 of these 48 patients, a positive PCR result was found at least once. Six of these patients developed a disseminated infection, defined as at least two consecutive positive samples, and three of these patients died (Table 5). One patient (UPN 497) died because of multifactorial complications not specifically caused by adenovirus. UPN 484 and UPN 510 had adenovirus-related complications as causes of death. Species C adenoviruses were detected in five of the six patients, and in one of the fatal cases HAdV 18 was the cause (Table 5). Patients with a viral load over 106 to 107 copies/ml may be considered at risk for severe complications.
|
View this table: [in a new window] |
TABLE 5. Prospective monitoring of adenovirus DNA in plasma samples of SCT patients in the first weeks after transplantation
|
![]() View larger version (14K): [in a new window] |
FIG. 1. Course of the adenovirus DNA load in four fatal cases of disseminated adenovirus infection. UPN 246 and 316 were from the retrospective study; UPN 484 and 510 were prospectively monitored. UPN 484 was treated with ribavirin starting in week 3, without a significant effect. The others were not treated.
|
|
|
|---|
In the present work, the development of a single-tube, internally controlled multiplex real-time PCR assay for monitoring the adenovirus DNA load in SCT patients is described. The heterogeneity of the adenovirus family substantially complicated the design of a quantitative PCR assay, as the possibilities for selecting general primers and probes are limited. Lion et al. (24) and Gu et al. (12) recently described a real-time PCR procedure using six different primer sets and probes for individual detection of all species. Heim et al. (14) described a broadly reactive adenovirus assay using a multiplex PCR.
Our assay has been designed to detect and quantify the serotypes involved in disseminated adenovirus disease. The HAdV DNA load measurement in the patients from a retrospective study generally confirmed the initial, semiquantitative PCR results (29). In one case, adenovirus type 41, which previously had been implicated in the fatal course of the disease, was found. However, the low viral loads observed in this patient make this conclusion questionable. In retrospect, Lion et al. (24) also detected a species F virus in a fatal case, and thus the role of species F adenoviruses will require further investigation.
The assay with the MGB probe enabled general detection and quantification of HAdV by using the same primers but without differentiation of species. The CT values with this probe are in general a little lower than when molecular beacons are used, but when the standard curve is also generated with the MGB probe, no significant differences in viral loads are obtained (data not shown). Therefore, this procedure can be used in laboratories where no multicolor real-time PCR detection system is available.
An internal control reaction is included in the multiplex PCR assay to monitor appropriate DNA isolation and check for inhibition. Although few samples were inhibitory (less than 0.5%), this control is important for the application of PCR to clinical samples, especially as the effect of drugs in the blood of these patients on the PCR is unknown.
In our and other pediatric patient populations, species C adenoviruses have been most frequently detected after transplantation. Interestingly, species C viruses have also been detected in the majority of tissue samples taken after tonsillectomy or adenoidectomy (10). Whether the increase in infections can be correlated with abandoning the practice of tonsillectomies is not clear yet. Species B adenoviruses, especially subtype 7, 11, 34, and 35 infections, have been described for transplant recipients as well and are the predominant species in solid organ transplant recipients (15, 18).
Clinical application of the assay for pediatric SCT recipients with fatal disseminated adenovirus infections showed adenovirus DNA loads of up to 1012 DNA copies/ml of plasma. In the nonfatal cases, the DNA load never exceeded 107 copies/ml except in a single case (UPN 485). These data suggest that patients with HAdV DNA loads of over 106 to 107 copies/ml in plasma or serum may be considered at risk for severe complications. Further studies applying preemptive treatment based on DNA loads should lead to a well-defined threshold for the start of therapeutic interventions. The potential role of the preemptive treatment strategy in these cases is one of the main issues in a multicenter study that has been recently initiated.
An important point of concern in defining thresholds for the initiation of treatment is the standardization of the HAdV DNA load results. The exact number of complete DNA copies in the HPLC-titrated standards used is unclear. Without standardized adenovirus stocks, the thresholds for preemptive therapy have to be defined locally. For general recommendations to identify patients at risk, true standardization by international reference materials has to become available, as recently has been established for nucleic acid testing of several viral targets in blood products (27, 28).
Ribavirin and cidofovir have been reported to be effective against adenovirus infections. However, in cases in which improvement of patients after treatment has been reported, no data on the viral load were available. Moreover, without any intervention, adenovirus viremia can be controlled as a result of immunological recovery of the patient. With quantitative HAdV DNA PCR, tools have become available to enable timely preemptive treatment initiation. Recently, our group has shown that ribavirin did not have convincing antiviral effects in SCT patients (21), and currently the efficacy of cidofovir is being evaluated in the aforementioned multicenter study.
The quantification of adenovirus DNA as described here will be used for the early detection of disseminated infections and as an objective instrument to evaluate therapeutic interventions. This approach should eventually lead to the definition of optimal treatment protocols for this serious complication of immunocompromised children.
We thank Sitha Scheltinga and Joyphi Thijssen for excellent technical assistance.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»