Abstract
The “gold standard” for the diagnosis of adenovirus (AV) infection is virus culture, which is rather time-consuming. Especially for immunocompromised patients, in whom severe infections with AV have been described, rapid diagnosis is important. Therefore, an internally controlled AV real-time PCR assay detecting all known human AV serotypes was developed. Primers were chosen from the hexon region, which is the most conserved region, and in order to cover all known serotypes, degenerate primers were used. The internal control (IC) DNA contained the same primer binding sites as the AV DNA control but had a shuffled probe region compared to the conserved 24-nucleotide consensus AV hexon probe region (the target). The IC DNA was added to the clinical sample in order to monitor extraction and PCR efficiency. The sensitivity and the linearity of the AV PCR were determined. For testing the specificity of this PCR assay for human AVs, a selection of 51 AV prototype strains and 66 patient samples positive for other DNA viruses were tested. Moreover, a comparison of the AV PCR method described herein with culture and antigen (Ag) detection was performed with a selection of 151 clinical samples. All 51 AV serotypes were detected in the selection of AV prototype strains. Concordant results from culture or Ag detection and PCR were found for 139 (92.1%) of 151 samples. In 12 cases (7.9%), PCR was positive while the culture was negative. In conclusion, a sensitive, internally controlled nonnested AV real-time PCR assay which is able to detect all known AV serotypes with higher sensitivity than a culture or Ag detection method was developed.
Today, 52 subtypes of adenoviruses (AV; family Adenoviridae, genus Mastadenovirus) are known to infect humans. They are transmitted by the fecal-oral route and the respiratory route and are associated with acute respiratory disease (accounting for 10% of febrile respiratory diseases in children), conjunctivitis, genitourinary infections, and infant gastroenteritis. AV have proved to be associated with the induction of malignant tumors in animals; however, this correlation has not been shown in humans. Human AV serotyping is based on resistance to neutralization by antisera to other AV serotypes. Of the 52 known serotypes, 51 have been sorted into six serogroups, based on their ability to agglutinate red blood cells from different species; serotype 52 has been found only recently. Most AV diseases are caused by a few serotypes (1 to 7) usually producing only mild infections in the immunocompetent host. The agent causing the infection can be isolated from stool samples during periods of no illness. This phenomenon hinders the establishment of a causal association of AV with disease and limits the significance of the diagnostic detection of these viruses. Based on virus culture studies, it is known that different serotypes of AV can cause different clinical syndromes; however, we may have underestimated the incidence of AV disease in some patient groups because some strains are difficult to culture (2, 15). Disseminated AV infections in immunocompromised individuals have been reported to yield high morbidity and mortality rates, especially among children (9, 10, 13, 18, 23, 26, 27). However, the epidemiology of AV disease is hampered by the facts that the sensitivity of AV culture is sometimes low and most AV PCR assays do not detect all known human AV types. These conclusions are supported by the findings in a recent publication of Casas et al. showing some new associations between specific clinical syndromes and various human AV serotypes (4). For instance, for the first time, a measles-like syndrome in persons previously vaccinated against measles was associated with AV serotypes 4 and 5.
Disseminating AV infections can be diagnosed by the culture of AV from specimens obtained from multiple body sites, but this method is time-consuming and not very sensitive (2, 12, 16). The detection of AV DNA in serum or plasma by PCR has been shown to predict disseminated AV infection very reliably (7, 19). In recent years, several sensitive AV PCR assays have been developed (1, 5, 11, 14, 19, 20, 24). However, these PCR assays do not detect all known human AV types, they are nested PCRs, or they are not all internally controlled. The aim of our study was to develop an internally controlled real-time PCR assay detecting all known human AV types.
MATERIALS AND METHODS
Viral strains.
A selection of AV prototype strains of 51 AV serotypes was kindly provided by J. C. de Jong, Erasmus MC, Rotterdam, The Netherlands.
Clinical samples.
One hundred fifty-one clinical samples from 96 patients suspected of having AV infection were tested in the AV PCR. This selection comprised 86 fecal samples (10 AV culture negative and AV antigen positive, 36 AV culture positive with an unknown AV antigen status, and 40 AV culture negative and AV antigen negative), 62 respiratory samples (23 AV culture positive and 39 AV culture negative), 2 skin swabs (1 AV culture positive and 1 AV culture negative), and 1 cerebrospinal fluid (CSF) specimen (AV culture negative).
For specificity testing, we used 41 EDTA plasma samples (2 PCR positive for BK virus, 6 PCR positive for cytomegalovirus [CMV], and 33 PCR positive for Epstein-Barr virus [EBV]), 4 urine samples PCR positive for BK virus, 13 fecal samples (10 PCR positive for CMV and 3 PCR positive for EBV), and 8 bronchoalveolar lavage (BAL) samples (5 PCR positive for CMV and 3 PCR positive for EBV).
Viral culture.
Clinical samples were cultured on human lung adenocarcinoma A549 cells, human diploid fibroblasts, tertiary monkey kidney cells, and Vero cells. The viral cultures were examined twice weekly for the appearance of an (AV-specific) cytopathological effect. The identification of the isolates was performed according to the cytopathological effect in unstained cultures or the staining seen after incubation with a specific monoclonal antibody (Dako, Glostrup, Denmark). In addition to virus culture, AV antigen detection (using an assay kit from Dako, Glostrup, Denmark) was performed with fecal samples in order to detect AV types which are difficult to culture (e.g., AV types 40 and 41). Viral titers of serotypes were determined by 50% tissue culture infective dose (TCID50) analysis according to the Reed-Muench method (21).
Primers and probe.
Primers from the hexon region of the AV were chosen. Forty-nine complete genomes representing all subgroups, as well as around 1,550 hexon genes of different serotypes and isolates, were analyzed in the Vector NTI Advance program (Invitrogen). The most conserved region of 103 bp was chosen as a target. In order to amplify all known types, the following degenerate primers were constructed: a forward primer, 5′-CAGGACGCCTCGGRGTAYCTSAG-3′, and a reverse primer, 5′-GGAGCCACVGTGGGRTT-3′ (where R is A or G, S is C or G, V is A, C, or G, and Y is C or T). The following 24-nucleotide consensus probe sequence was chosen: 5′-CCGGGTCTGGTGCAGTTTGCCCGC-3′. Both the forward (n = 8) and reverse (n = 6) primers had maximum redundancy.
Alignments of sequences.
All known complete genome sequences of AV were aligned using Vector NTI and ClustalW.
Construction of the AV-containing plasmid.
AV DNA was purified from 200 μl of AV type 2 (AV2) stock (1 in 1,000 dilution) in lysis buffer as described in “DNA purification” below. Amplification was performed with the following nondegenerate target primers: forward, 5′-CAGGACGCCTCGGAGTACCTGAG-3′, and reverse, 5′-GGAGCCACCGTGGGGTT-3′. The 103-bp amplicon was cloned into a PCRII-TOPO plasmid according to the instructions of the manufacturer (Invitrogen). Verification of the AV-containing plasmid was performed by sequencing. The concentration of AV DNA from the AV-containing plasmid was determined by evaluating the optical density at 260 nm, and serial dilutions of AV DNA were used to determine the sensitivity of the AV PCR.
Construction of the IC-containing plasmid.
We designed two oligonucleotides (linkers) for the construction of internal control (IC) DNA, which were synthesized by Applied Biosystems: adeno hexon linker 1 (5′-CAGGACGCCTCGGAGTACCTGAGCCGATGTGTCCGCCGTGGTCCCCTGGACCGAGACGTACTT-3′) and adeno hexon linker 2 (5′-GGAGCCACCGTGGGGTTTCTAAACTTGTTATTCAGGCTGAAGTACGTCTCGGTCCAGGGGACCACGG-3′). These two linkers, which together represent the same 103-bp hexon region as that present in the in vitro AV DNA control, overlapped over a stretch of 28 nucleotides (underlined) and contained the same primer binding sites as the in vitro AV DNA control (doubly underlined) but with a shuffled probe region (bold). The IC probe region allows discrimination between AV and IC DNA amplicons during amplification and detection. The IC DNA control was constructed by the hybridization and elongation of 1 ng of linker 1 and 1 ng of linker 2 in a mixture of 2.5 U of AmpliTaq gold, 5 μg of bovine serum albumin, 1× PCR II buffer, deoxynucleoside triphosphates at a concentration of 200 μM each, and 3 mM MgCl2. The mixture was incubated for 10 min at 95°C, 5 min at 55°C, and 10 min at 72°C. The resulting hybrid was subsequently amplified with the nondegenerate target primers and cloned into a PCRII-TOPO plasmid according to the instructions of the manufacturer (Invitrogen). Verification of the IC-containing plasmid was performed by sequencing. The concentration of DNA was determined by evaluating the optical density at 260 nm.
DNA purification.
The 51 prototypes were isolated alternately with negative controls by using the MagNA Pure (MP) system (Roche Diagnostics, Penzberg, Germany) as follows: 5 μl of each prototype together with 104 IC DNA copies was mixed with 350 μl of MP lysis buffer. For isolation from the clinical samples, we used 200 μl of throat fluid, sputum, or plasma and 350 μl of MP lysis buffer or 50 μl of fecal material and 500 μl of MP lysis buffer. These mixtures were then subjected to a vortex in an Eppendorf tube, left for 10 min at room temperature (prelysis step), and subsequently centrifuged for 2 min at 13,000 rpm in an Eppendorf centrifuge. Thereafter, 490 μl of supernatant was transferred into an MP sample cartridge together with 10,000 copies of IC DNA. Isolation was then performed with the MP system according to the protocol of the manufacturer (Roche Diagnostics, Penzberg, Germany) by using the total nucleic acid kit. The DNA was finally eluted in 100 μl of MP elution buffer.
Competitive TaqMan PCR.
Ten microliters of each eluate, containing 1,000 copies of IC DNA, was used for the TaqMan PCR. The final PCR mixture (25 μl) contained 12.5 μl of TaqMan Universal PCR master mix (ABI), 900 nM forward primer, 900 nM reverse primer, 200 nM target probe, 200 nM IC probe, and 400 ng of α-casein/μl (2). PCR was performed with an ABI Prism 7000 sequence detection system as follows: 2 min at 50°C and 10 min at 95°C, followed by 45 cycles consisting of 15 s at 95°C and 1 min at 60°C. A signal was considered to be relevant if a logarithmic curve was visible above the threshold for the target and/or the IC.
Serial dilutions.
Twelve (twofold) serial dilutions of plasmid containing AV2 hexon DNA (AV2-plasmid DNA) and IC DNA in Tris-EDTA (pH 8.0) with 20 ng/μl of calf thymus DNA (Sigma, The Netherlands) in a background of AV-negative throat fluid were made and tested by PCR after the extraction of DNA with the MP system in order to test the lower limit of detection (LLOD) of the AV PCR. Furthermore, 12 (10-fold) serial dilutions of AV2-plasmid DNA in a background of AV-negative plasma, with 104 copies of IC DNA in each dilution, were made in order to investigate whether there was an effect of competition from the IC on the linearity and the LLOD of the assay.
RESULTS
Determination of the LLOD of the AV real-time PCR assay.
To determine the LLOD of the AV real-time PCR assay, we spiked 200-μl aliquots of AV-negative throat fluid with decreasing amounts of AV2-plasmid DNA, as well as IC DNA, before DNA purification by the MP system. DNA was eluted in 100 μl, and 10 μl was used for real-time PCR. Twelve series of twofold dilutions of AV DNA and IC DNA were tested. As shown in Table 1, limiting dilutions of the AV DNA revealed a detection limit of 8 AV DNA copies in the PCR mixture, with a 50% hit rate (6 of 12 runs), resulting in an analytical detection limit of 400 copies/ml. Limiting dilutions of the IC DNA revealed a detection limit of 16 IC DNA copies in the PCR mixture, with a 42% hit rate (5 of 12 runs), resulting in an analytical detection limit of around 103 copies/ml (Table 1).
TABLE 1.
LLODs for plasmids containing part of the AV2 hexon gene DNA and IC DNAa
| Sample | No. of DNA copies in PCR mixtureb | Proportion positive (%) |
|---|---|---|
| AV2 hexon gene DNA | 250 | 12/12 (100) |
| 125 | 12/12 (100) | |
| 62 | 12/12 (100) | |
| 31 | 10/12 (83.3) | |
| 16 | 10/12 (83.3) | |
| 8 | 6/12 (50.0) | |
| 4 | 1/12 (8.3) | |
| 0 | 0/12 (0) | |
| IC DNA | ||
| 250 | 12/12 (100) | |
| 125 | 12/12 (100) | |
| 62 | 11/12 (91.7) | |
| 31 | 10/12 (83.3) | |
| 16 | 5/12 (41.7) | |
| 8 | 3/12 (25.0) | |
| 4 | 2/12 (16.7) | |
| 0 | 0/12 (0) |
The slope (resulting from a comparison between the number of copies in the PCR mixture and the hit rate) for AV2 hexon gene DNA was 3.68 ± 0.13 (mean ± standard error of the mean), and that for IC DNA was 3.79 ± 0.009.
With 100% efficiency in extraction.
Determination of the linearity of the AV real-time PCR assay with a constant amount of IC.
To determine the linearity of the AV real-time PCR assay, we spiked 200-μl samples of AV-negative plasma with decreasing amounts of AV2-plasmid DNA. Before DNA purification by the MP system, every sample was spiked with 104 copies of IC DNA. DNA was eluted in 100 μl, and 10 μl of extracted DNA was used for real-time PCR, resulting in 103 copies of IC DNA per sample. Twelve series of 10-fold dilutions of AV DNA were tested. The limit of detection of AV DNA in this setting was 100 AV DNA copies in the PCR mixture, with a 100% hit rate (12 of 12 runs), resulting in a 100% quantitative limit of detection of 5 × 103 copies of AV DNA per ml (Table 2). The quantitative AV PCR has a linear dynamic range between 5 × 103 and 5 × 108 copies/ml, with a regression coefficient of 0.991 (Fig. 1).
TABLE 2.
Linearity of extraction results for plasmid containing part of the AV2 hexon gene DNA with 104 copies of IC DNA
| No. of AV DNA copies in PCR mixture | No. of samples positive for AV DNA/total no. of samples (%) | Mean CTa for AV DNA | No. of samples positive for IC DNA/total no. of samples (%) | Mean CTa for IC DNA |
|---|---|---|---|---|
| 107 | 12/12 (100) | 18.69 | 0/12 (0) | NDb |
| 106 | 12/12 (100) | 22.58 | 5/12 (41.7) | 40.78 |
| 105 | 12/12 (100) | 26.22 | 12/12 (100) | 38.54 |
| 104 | 12/12 (100) | 29.92 | 12/12 (100) | 34.70 |
| 103 | 12/12 (100) | 33.28 | 12/12 (100) | 34.58 |
| 102 | 12/12 (100) | 37.11 | 12/12 (100) | 34.07 |
| 101 | 1/12 (8.3) | 44.19 | 12/12 (100) | 34.43 |
| 0 | 0/12 (0) | NDb | 12/12 (100) | 34.33 |
The mean cycle threshold (CT) is the average CT value of 12 independent measurements.
ND, not detectable.
FIG. 1.
Standard curve for plasmid DNA containing part of the AV2 hexon gene. Twelve series of 10-fold dilutions of extracted plasmid DNA containing part of the AV2 hexon gene, each with a constant level of 104 copies of IC DNA per dilution, were tested by PCR, resulting in a dynamic range of 5 × 103 to 5 × 108 copies/ml, with a regression coefficient of 0.991.
Determination of the efficiency of the AV real-time PCR assay in comparison to the TCID50 method for representative serotypes belonging to the six subgroups.
In order to investigate if the redundancy of the primers influenced the efficiency of amplification of prototypes representing all six subgroups, limiting dilutions of prototype 2 (subgroup C), prototype 4 (subgroup E), prototype 7 (subgroup B), prototype 8 (subgroup D), prototype 31 (subgroup A), and prototype 41 (subgroup F) strains were performed. No differences in amplification efficiency among the six subgroups were found, and overall the PCR was more sensitive than the TCID50 titer determination method (Table 3). In addition, an alignment of the deduced amplicons of all known complete AV genomes is shown in Fig. 2. As the assay was intended to amplify all known serotypes, the most conserved (hexon) region was chosen. Degenerate primers were used, without the resulting amplicons’ being significant enough to serve as a template for genotyping differences, because the overall level of homology of the amplicons among the serotypes within each of the six subgroups is high.
TABLE 3.
Comparison of results from PCR and TCID50 methods
| Virus subgroup or DNA source | Serotype | Log no. of copies/ml by PCR | R2 value | Log no. of copies/ml by TCID50 method |
|---|---|---|---|---|
| Virus subgroups | ||||
| A | 31 | 8.5 | 0.9849 | 6.6 |
| B | 7 | 9.0 | 0.9991 | 7.1 |
| C | 2 | 9.3 | 0.9950 | 8.5 |
| D | 8 | 8.5 | 0.9784 | 3.9 |
| E | 4 | 9.2 | 0.9977 | 8.4 |
| F | 41 | 9.1 | 0.9869 | 2.0 |
| Plasmid | 2 | 7.1 | 0.9996 | NAb |
The sensitivity of PCR for different serotypes was determined by using limiting dilutions and comparing the results to TCID50 titers expressed as log values. One serotype from each subgroup was used in both PCR and TCID50 analyses.
NA, not applicable.
FIG. 2.
Alignments of deduced amplicon sequences from serogroups A to F. Alignments were made using Vector NTI. Primer binding sites are indicated in yellow. The probe binding site is indicated in blue. Redundancies in the primer and probe binding sites are marked in red, whereas dashes represent identical nucleotides. Forward primer sequence, CAGGACGCCTCGGRGTAYCTSAG; probe sequence, CCGGTCTGGTGCAATTCGCCCGC; and reverse primer sequence, TTRGGGTGVCACCGAGG (where R is A or G, S is G or C, V is A, C, or G, Y is C or T, and B is T, G, or C.)
Testing of the specificity of the AV real-time PCR assay.
The specificity of the real-time AV PCR assay was examined by testing 51 AV serotypes. The extraction of every single AV prototype was done alternately with negative extraction controls, and all samples contained IC DNA. Furthermore, 6 EDTA plasma samples PCR positive for CMV, 33 plasma samples PCR positive for EBV, 2 plasma samples PCR positive for BK virus, 4 urine samples PCR positive for BK virus, 3 fecal samples PCR positive for EBV, 10 fecal samples PCR positive for CMV, 3 BAL samples PCR positive for EBV, and 5 BAL samples PCR positive for CMV were tested by the AV real-time PCR. All 51 AV prototype strain samples were found to be positive for AV (Table 4) (the recently found serotype 52 [17] was checked in silico, and a 100% match with the AV-specific primer and probe sequences described herein was found); 57 of the 61 clinical samples tested negative for AV, whereas 4 of 61 samples tested positive for AV. These samples were all obtained from hematology patients, and AV positivity was confirmed by sequencing (results not shown). Negative results were not due to the presence of inhibitory substances since the signals for the coextracted IC DNA were positive in all cases (Table 5).
TABLE 4.
Sensitivity of AV PCR for prototype strains of AVa
| Subgroup (serotype[s]) | Source | Result | Target CT range |
|---|---|---|---|
| A (types 12, 18, 31) | Cell culture | Positive | 19.97-25.47 |
| B (types 3, 7, 11, 14, 16, 21, 34, 35, 50) | Cell culture | Positive | 22.57-37.75 |
| C (types 1, 2, 5, 6) | Cell culture | Positive | 19.13-21.53 |
| D (types 8-10, 13, 15, 17, 19, 20, 22-30, 32, 33, 36-39, 42-49, 51) | Cell culture | Positive | 13.32-20.69 |
| E (type 4) | Cell culture | Positive | 17.35 |
| F (types 40, 41) | Cell culture | Positive | 15.88-17.24 |
Results were confirmed by sequencing.
TABLE 5.
Specificity of AV PCR
| Virus | Source (no. of samples) | Result | Target CT | IC CT value or range |
|---|---|---|---|---|
| CMV | EDTA plasma (6) | Negative | NDa | 34.02-35.01 |
| BAL fluid (5) | Negative | ND | 33.85-34.88 | |
| Feces (10) | Negative | ND | 30.05-35.06 | |
| EBV | EDTA plasma (32) | Negative | ND | 32.40-35.90 |
| EDTA plasma (1) | Positiveb | 38.11 | 33.69 | |
| BAL fluid (2) | Negative | ND | 33.80-34.66 | |
| BAL fluid (1) | Positiveb | 30.57 | 42.25 | |
| Feces (2) | Negative | ND | 31.27-31.71 | |
| Feces (1) | Positiveb | 34.5 | 32.15 | |
| BK virus | EDTA plasma (1) | Negative | ND | 32.66 |
| EDTA plasma (1) | Positiveb | 30.83 | 31.91 | |
| Urine (4) | Negative | ND | 31.79-32.56 |
ND, not detectable.
Confirmed by sequencing.
Comparison of virus culture with AV real-time PCR for clinical specimens.
A panel of 151 clinical samples was collected for the evaluation of the AV real time PCR. Culture results for these samples were already available. The panel comprised 62 respiratory specimens (throat fluid, sputum, or BAL specimens), of which 23 were AV culture positive (1 of the 23 was also herpes simplex virus positive) and 39 were AV culture negative. Twenty-one of these 39 specimens were completely culture negative, and the other samples were culture positive for other viruses (enterovirus [EV], CMV, respiratory syncytial virus [RSV], or influenza virus A). Twenty-two of 39 AV culture-negative respiratory samples were from patients who had an AV-positive sample obtained from another body site and/or on another sampling date.
Furthermore, 86 fecal samples were included in the analysis; 36 were AV culture positive, 10 were AV culture negative but AV antigen positive by an enzyme-linked immunosorbent assay (ELISA), and 40 were AV culture and AV antigen negative. Twenty-eight of 40 specimens were completely culture negative, 10 of 40 were EV culture positive, 1 of 40 was rotavirus positive by an ELISA, and 1 of 40 was positive for Clostridium toxin. Two skin swabs were included in the selection, of which one was AV culture positive and one was EV culture positive; one CSF specimen, which was completely culture negative, was also included.
As depicted in Table 6, agreement between the results of virus culture (and antigen detection) and AV real-time PCR was found for 139 of 151 clinical samples (92.1%). Concordant negative results for 31 respiratory samples, 36 fecal samples, 1 skin swab, and 1 CSF specimen (45.7%) were found. Concordant positive results for 23 respiratory samples, 36 fecal samples, and 1 skin swab (39.7%) were found. Discordant results for 12 clinical samples (7.9%) were found. Eight respiratory specimens and four fecal samples (7.9%) were negative by AV culture and positive by the AV real-time PCR. Four of the 12 discrepancies were for samples from AV-infected patients (who had previous samples or other types of specimens that were culture positive); 7 of 12 were for samples that were culture or ELISA positive for another viral pathogen (CMV, rotavirus, or RSV), and 1 of 12 was for a sample positive for Clostridium toxin. In only one case (that of a sputum sample), the patient was not known to have an AV infection detected in other samples or on other test dates and no other pathogen was cultured (Table 6).
TABLE 6.
Comparison of results of AV PCR and virus culture or antigen detection for different clinical specimens
| Specimen | n | No. PCR and culture positive | No. PCR positive, culture negative, and ELISA positivea | No. PCR positive and culture negativeb | No. PCR negative and culture positive | No. PCR and culture negative |
|---|---|---|---|---|---|---|
| Feces | 86 | 36 | 10 | 4 | 0 | 36 |
| Respiratory specimen | 62 | 23 | 0 | 8 | 0 | 31 |
| Skin swab | 2 | 1 | 0 | 0 | 0 | 1 |
| CSF | 1 | 0 | 0 | 0 | 0 | 1 |
| Total (%) | 151 | 60 (39.7) | 10 (6.6) | 12 (7.9) | 0 (0) | 69 (45.7) |
PCR-positive, culture-negative, ELISA-positive samples were fecal samples with AV types which are probably not culturable on routine cells. All had high AV DNA loads.
Of the 12 PCR-positive, culture-negative samples, 4 were from AV-infected patients (with an earlier sample or different type of specimen culture positive), 7 were culture or ELISA positive for another viral pathogen (CMV, rotavirus, or RSV), and 1 was positive for Clostridium toxin.
DISCUSSION
The isolation of AV in cell culture, especially from respiratory materials and feces, is still regarded as the diagnostic “gold standard.” The disadvantages of culture are, however, obvious—for example, labor-intensiveness, delays of days to sometimes weeks to obtain a positive result, and false negativity. The latter can occur since some AV types (e.g., AV types 40 and 41) are difficult to culture (8, 15). The recognition of invasive AV disease has been facilitated by the development of AV PCR assays (14, 22, 24, 25). Only recently, we have learned that invasive AV disease can be an important complication in stem cell recipients, especially children. However, the studies have been performed with PCR assays which detect only a subset of AV types. The importance of other, less commonly isolated AV types is not known. Furthermore, other groups of patients need to be tested for all known AV serotypes, for example, solid-organ transplant recipients and other immunocompromised patients, including adults as well as children.
Various AV PCR assays have been developed up to now; however, none could detect all known AV serotypes in one assay together with an IC in a real-time setting. Various assays were developed which were able to detect a subset of AV types, with or without an IC and either in a real-time setting or not (1, 5, 11, 19, 24). Lion et al. (19) published a description of an AV PCR approach which is able to detect all 51 serotypes; however, six separate assays (six primer-probe combinations) are needed for the detection of all serotypes. Heim et al. (14) and Sarantis et al. (22) describe a TaqMan PCR method which is able to detect all 51 serotypes; however, in this assay, no IC is used. ICs are essential in diagnostic assays, because false-negative results or invalid results can be ruled out (3). Moreover, the diagnosis of AV infections with one set of primers and two probes for the detection of all known AV types and with an IC in a real-time PCR is preferable to the other approaches and to our knowledge not previously documented.
The real-time AV PCR assay described in the present paper is sensitive, has a broad linear range, and is specific. The sensitivity of the PCR assay was evaluated with limiting dilutions of both AV2-plasmid DNA and IC DNA. AV2-plasmid DNA and IC DNA in a background of AV-negative throat fluid were separately extracted by the MP system. Poisson statistics predict that 63% of the reactions will be positive with a single copy of DNA in the PCR mixture (6). The differences observed in the detection rate between AV2-plasmid DNA and IC DNA are therefore probably within the normal test variation.
The AV IC DNA was constructed from an AV2 stock. We evaluated the sensitivity of the PCR for AV serotypes belonging to the six subgroups in serial dilutions in comparison to that of the TCID50 method of determining titers, and no significant differences were found. Table 3 shows results for representative serotypes from the six subgroups in comparison to TCID50 results. In all cases, the PCR was more sensitive. Moreover, the titration of other serotypes showed similar trends, with comparable R2 values. This finding suggests comparable sensitivities of the PCR for the various subgroups and serotypes.
The PCR was tested against a selection of all known 51 prototype strains with alternating negative controls: 6 EDTA plasma samples PCR positive for CMV, 33 plasma samples PCR positive for EBV, 2 plasma samples PCR positive for BK virus, 4 urine samples PCR positive for BK virus, 3 fecal samples PCR positive for EBV, 10 fecal samples PCR positive for CMV, 3 BAL samples PCR positive for EBV, 5 BAL samples PCR positive for CMV, and 151 clinical samples with known culture results. The results for the panel of samples with the 51 prototype strains showed that our PCR detected them all, whereas serotype 52 (17) showed 100% homology to our primer and probe sequences, making it likely to be detected by the AV PCR assay described herein. The specificity of our AV real-time PCR assay was tested with 66 samples positive for other DNA viruses, and 62 of 66 (94%) were AV PCR negative. The four exceptions came from patients with a hematological disease, and other materials from these patients were also positive in the viral culture. Moreover, AV infection in these particular samples was confirmed by sequencing.
Comparisons with virus culture showed good concordance, and in 12 of 151 cases (7.9%), the PCR was positive while virus culture was negative. In 4 of these 12 cases, AV was cultured from another type of specimen from the same patient, earlier or later in the course of the disease. In these cases, it is conceivable that, due to better sensitivity, the PCR detected AV while the virus culture was negative. In 6 of 12 cases, another viral pathogen (CMV or RSV) was cultured; in 1 of 12, the sample was positive for rotavirus antigen; and in 1 of 12, the cells were destroyed by Clostridium toxin. For these eight samples, it is possible that the culture for AV was less sensitive than the AV PCR assay due to the presence of other pathogens (resulting in overgrowth by another pathogen and/or low AV loads in patient samples because of interference).
In this paper, we present a newly developed internally controlled AV real-time PCR assay which is sensitive when tested on serial dilutions as well as when tested on clinical samples. It is specific and able to detect all known AV serotypes with one primer pair-probe set. The assay is easy to use in diagnostic as well as in research settings, because only one test run per sample is required. Since the spectrum of AV disease is not fully known, our AV real-time PCR may be an important tool in further research and diagnostic protocols.
Acknowledgments
We thank J. C. de Jong, Erasmus MC, Rotterdam, The Netherlands, for kindly providing the AV prototype strains and peer reviewing and Nicholas Griffin for critically reviewing and reading the manuscript.
Footnotes
Published ahead of print on 15 October 2008.
REFERENCES
- 1.Avalón, A., P. Pérez, J. C. Aguilar, R. O. De Lejarazu, and J. E. Echevarría. 2001. Rapid and sensitive diagnosis of human adenovirus infections by a generic polymerase chain reaction. J. Virol. Methods 92113-120. [DOI] [PubMed] [Google Scholar]
- 2.Baldwin, A., H. Kingman, M. Darville, A. B. M. Foot, D. Grier, J. M. Cornish, N. Goulden, A. Oakhill, D. H. Pamphilon, C. G. Steward, and D. I. Marks. 2000. Outcome and clinical course of 100 patients with adenovirus infection following bone marrow transplantation. Bone Marrow Transplant. 261333-1338. [DOI] [PubMed] [Google Scholar]
- 3.Beld, M., R. Minnaar, J. Weel, C. Sol, M. Damen, H. van der Avoort, P. Wertheim-van Dillen, A. van Breda, and R. Boom. 2004. Highly sensitive assay for detection of enterovirus in clinical specimens by reverse transcription-PCR with an armored RNA internal control. J. Clin. Microbiol. 423059-3064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Casas, I., A. Avellon, M. Mosquera, O. Jabado, J. E. Echevarria, R. H. Campos, M. Rewers, P. Perez-Breña, W. I. Lipkin, and G. Palacios. 2005. Molecular identification of adenoviruses in clinical samples by analyzing a partial hexon genomic region. J. Clin. Microbiol. 436176-6182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Claas, E. C., M. W. Schilham, C. S. de Brouwer, P. Hubacek, M. Echavarria, A. C. Lankester, M. J. van Tol, and A. C. Kroes. 2005. Internally controlled real-time PCR monitoring of adenovirus DNA load in serum or plasma of transplant recipients. J. Clin. Microbiol. 431738-1744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Diaco, R. 1995. Practical considerations for the design of quantitative PCR assays, p. 84-108. In M. A. Innis, D. H. Gelfland, and J. J. Sninsky (ed.), PCR strategies. Academic Press, Inc., New York, NY.
- 7.Echavarria, M., M. Forman, M. J. van Tol, J. M. Vossen, P. Charache, and A. C. Kroes. 2001. Prediction of severe disseminated adenovirus infection by serum PCR. Lancet 358384-385. [DOI] [PubMed] [Google Scholar]
- 8.Echavarria, M., S. C. Ray, R. Ambinder, J. S. Dumler, and P. Charache. 1999. PCR detection of adenovirus in a bone marrow transplant recipient: hemorrhagic cystitis as a presenting manifestation of disseminated disease. J. Clin. Microbiol. 37686-689. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Echavarria, M., S. A. Kolavic, S. Cersovsky, F. Mitchell, J. L. Sanchez, C. Polyak, B. L. Innis, and L. N. Binn. 2000. Detection of adenoviruses (AdV) in culture-negative environmental samples by PCR during an AdV-associated respiratory disease outbreak. J. Clin. Microbiol. 382982-2984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Flomenberg, P., J. Babbitt, W. R. Drobyski, R. C. Ash, D. R. Carrigan, G. V. Sedmak, T. McAuliffe, B. Camitta, M. M. Horowitz, N. Bunin, and J. T. Casper. 1994. Increasing incidence of adenovirus disease in bone marrow transplant recipients. J. Infect. Dis. 169775-781. [DOI] [PubMed] [Google Scholar]
- 11.Gu, Z., S. W. Belzer, C. S. Gibson, M. J. Bankowski, and R. T. Hayden. 2003. Multiplexed, real-time PCR for quantitative detection of human adenovirus. J. Clin. Microbiol. 414636-4641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hale, G. A., H. E. Heslop, R. A. Krance, M. A. Brenner, D. Jayawardene, D. K. Srivastava, and C. C. Patrick. 1999. Adenovirus infection after pediatric bone marrow transplantation. Bone Marrow Transplant. 23277-282. [DOI] [PubMed] [Google Scholar]
- 13.Heemskerk, B., A. C. Lankester, T. van Vreeswijk, M. F. C. Beersma, E. C. J. Claas, L. A. Veltrop-Duits, A. C. M. Kroes, J. M. J. J. Vossen, M. W. Schilham, and M. J. D. van Tol. 2005. Immune reconstitution and clearance of human adenovirus viremia in pediatric stem-cell recipients. J. Infect. Dis. 191520-530. [DOI] [PubMed] [Google Scholar]
- 14.Heim, A., C. Ebnet, G. Harste, and P. Pring-Akerblom. 2003. Rapid and quantitative detection of human adenovirus DNA by real-time PCR. J. Med. Virol. 70228-239. [DOI] [PubMed] [Google Scholar]
- 15.Horwitz, M. S. 2001. Adenoviruses, p. 2301-2326. In D. M. Knipe and P. M. Howley (ed.), Fields virology, 4th ed. Lippincott Williams and Wilkins, Philadelphia, PA.
- 16.Howard, D. S., G. L. Phillips II, D. E. Reece, R. K. Munn, J. Henslee-Downey, M. Pittard, M. Barker, and C. Pomeroy. 1999. Adenovirus infections in hematopoietic stem cell transplant recipients. Clin. Infect. Dis. 291494-1501. [DOI] [PubMed] [Google Scholar]
- 17.Jones, M. S., II, B. Harrach, R. D. Ganac, M. M. A. Gozum, W. P. dela Cruz, B. Riedel, C. Pan, E. L. Delwart, and D. P. Schnurr. 2007. New adenovirus species found in a patient presenting with gastroenteritis. J. Virol. 815978-5984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Leruez-Ville, M., V. Minard, F. Lacaille, A. Buzyn, E. Abachin, S. Blanche, F. Freymuth, and C. Rouzioux. 2004. Real-time blood plasma polymerase chain reaction for management of disseminated adenovirus infection. Clin. Infect. Dis. 3845-52. [DOI] [PubMed] [Google Scholar]
- 19.Lion, T., R. Baumgartinger, F. Watzinger, S. Matthes-Martin, M. Suda, S. Preuner, B. Futterknecht, A. Lawitschka, C. Peters, U. Potschger, and H. Gadner. 2003. Molecular monitoring of adenovirus in peripheral blood after allogeneic bone marrow transplantation permits early diagnosis of disseminated disease. Blood 1021114-1120. [DOI] [PubMed] [Google Scholar]
- 20.Pehler-Harrington, K., M. Khanna, C. R. Waters, and K. J. Henrickson. 2004. Rapid detection and identification of human adenovirus species by Adenoplex, a multiplex PCR-enzyme hybridization assay. J. Clin. Microbiol. 424072-4076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Reed, L. J., and H. Muench. 1938. A simple method of estimating fifty percent endpoints. Am. J. Hyg. 27493-497. [Google Scholar]
- 22.Sarantis, H., G. Johnson, M. Brown, M. Petric, and R. Tellier. 2004. Comprehensive detection and serotyping of human adenoviruses by PCR and sequencing. J. Clin. Microbiol. 423963-3969. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Schilham, M. W., E. C. Claas, W. van Zaane, B. Heemskerk, J. M. Vossen, A. C. Lankester, R. E. Toes, M. Echavarria, A. C. Kroes, and M. J. van Tol. 2002. High levels of adenovirus DNA in serum correlate with fatal outcome of adenovirus infection in children after allogeneic stem-cell transplantation. Clin. Infect. Dis. 35526-532. [DOI] [PubMed] [Google Scholar]
- 24.Seidemann, K., A. Heim, E. D. Pfister, H. Köditz, A. Beilken, A. Sander, M. Melter, K. Sykora, M. Sasse, and A. Wessel. 2004. Monitoring of adenovirus infection in pediatric transplant recipients by quantitative PCR: report of six cases and review of the literature. Am. J. Transplant. 42102-2108. [DOI] [PubMed] [Google Scholar]
- 25.Vabret, A., S. Gouarin, M. Joannes, C. Barranger, J. Petitjean, S. Corbet, J. Brouard, F. Lafay, J. F. Duhamel, B. Guillois, and F. Freymuth. 2004. Development of a PCR- and hybridization-based assay (PCR Adenovirus Consensus) for the detection and the species identification of adenoviruses in respiratory specimens. J. Clin. Virol. 31116-122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.van Tol, M. J. D., E. C. J. Claas, B. Heemskerk, L. A. Veltrop-Duits, C. S. de Brouwer, T. van Vreeswijk, C. C. Sombroek, A. C. M. Kroes, M. F. C. Beersma, E. P. A. de Klerk, R. M. Egeler, A. C. Lankester, and M. W. Schilham. 2005. Adenovirus infection in children after allogeneic stem cell transplantation: diagnosis, treatment and immunity. Bone Marrow Transplant. 35S73-S76. [DOI] [PubMed] [Google Scholar]
- 27.Walls, T., A. G. Shankar, and D. Shingadia. 2003. Adenovirus: an increasingly important pathogen in paediatric bone marrow transplant patients. Lancet Infect. Dis. 379-86. [DOI] [PubMed] [Google Scholar]