|Home | About | Journals | Submit | Contact Us | Français|
Human adenoviruses (AdV) have been implicated in a wide variety of diseases and are ubiquitous in populations worldwide. These agents are of concern particularly in immunocompromised patients, children, and military recruits, resulting in severe disease or death. Clinical diagnosis of AdV is usually achieved through routine viral cell culture, which can take weeks for results. Immunofluorescence and enzyme-linked immunosorbent assay-based techniques are more timely but lack sensitivity. The ability to distinguish between the six different AdV species (A to F) is diagnostically relevant, as infections with specific AdV species are often associated with unique clinical outcomes and epidemiological features. Therefore, we developed a multiplex PCR-enzyme hybridization assay, the Adenoplex, using primers to the fiber gene that can simultaneously detect all six AdV species A through F in a single test. The limit of detection (LOD) based on the viral 50% tissue culture infective dose/ml for AdV A, B, C, D, E, and F was 10−2, 10−1, 10−1, 10−2, 10−1, and 10−2, respectively. Similarly, the LOD for the six DNA controls ranged from 102 to 103 copies/ml. Twelve common respiratory pathogens were tested with the Adenoplex, and no cross-reactivity was observed. We also validated our assay using clinical specimens spiked with different concentrations of AdV strains of each species type and tested by multiplex PCR and culture. The results demonstrated an overall sensitivity and specificity of Adenoplex of 100%. This assay can be completed in as few as 5 h and provides a rapid, specific, and sensitive method to detect and subtype AdV species A through F.
Human adenoviruses (AdV) cause a variety of diseases and are prevalent throughout the world. Common clinical manifestations resulting from AdV infection include pneumonia, cystitis, conjunctivitis, diarrhea, hepatitis, myocarditis, and encephalitis (10). In the general population, AdV often cause mild or self-limiting disease; however, severe disseminated disease can occur in immunocompromised individuals and can be fatal. Pediatric bone marrow transplant patients are at increased risk for AdV infection and have high mortality rates (22). Discontinuation of vaccinating U.S. military trainees against AdV has resulted in a resurgence of respiratory disease epidemics in this population, leading to increased AdV morbidity (12, 15).
There are currently 51 recognized AdV serotypes that have been grouped into six different species (formerly subgenera), A to F, based on their physiochemical, biological, and genetic properties (3, 21). Identification of AdV species types can be particularly useful to clinicians, as specific species can cause infections with unique clinical outcomes and epidemiological features. For example, the ability to distinguish between group D AdV, which cause severe and highly contagious keratoconjunctivitis, and group B and E AdV, which cause mild ocular infections, may be of great clinical importance. Also, species typing may help identify disease from asymptomatic chronic infection and also nosocomial infections.
The “gold standard” of diagnosis for AdV infection has been viral culture. Although culture provides a sensitive method for diagnosis, it can take as long as 3 weeks to achieve results. Immunofluorescence and other immunodiagnostic assays using direct antigen detection are more rapid than culture, but they lack sensitivity. Recently, PCR-based assays have been employed in clinical practice for the detection of AdV and have proven to be comparable or better than the classic methods (1, 4, 5, 8, 11, 13, 17, 18, 23). However, many of these assays do not allow for AdV species typing in a rapid and efficient manner. Thus, we developed a multiplex PCR-enzyme hybridization assay, the Adenoplex, for rapid, simultaneous detection and identification of AdV species in a single test.
AdV strains (serotypes 1 through 9, 11, 12, 16, 19, 21, 30, 31, 34, 35, 37, 40, 41, 48, and 49), human parainfluenza viruses 1, 2, and 3, influenza viruses A and B, respiratory syncytial viruses A and B, Bordetella pertussis, Chlamydia pneumoniae, Legionella micdadei, Legionella pneumophila, and Mycoplasma pneumoniae strains used in this study were obtained from the American Type Culture Collection (Manassas, Va.).
A total of 150 nasopharyngeal (NP) swab specimens, which were previously determined to be AdV negative, were used in our study. We spiked 20 each (n = 120) with AdV A, B, C, D, E, and F and left 30 unspiked as negative controls. Fifty percent of the specimens were spiked at 103 50% tissue culture infective doses (TCID50)/ml, and the other 50% were spiked at 102 TCID50/ml. A total sample volume of 2.0 ml was split equally for blind testing by either culturing (City of Milwaukee Health Department, Milwaukee, Wis.) or by the Adenoplex assay. For discrepant analysis, frozen DNA from the specimens that were culture negative and Adenoplex positive was sent to another commercial laboratory for testing using their AdV PCR-based assay.
A volume of 1.0 ml of clinical specimen was split into two 0.5-ml aliquots, and each aliquot was inoculated onto two cell types. Specimens spiked with species A to E were inoculated onto HEp2 and Caco2 cells, and specimens spiked with species F were inoculated onto RMK-1 and Caco2 cells. Cells were cultured for 4 to 14 days and observed for cytopathic effect (CPE). Cultures that did not have adequate CPE development for quantification of virus after 14 days were then fixed with methanol, washed, and stained using the Light Diagnostics AdV direct immunofluorescence assay (Chemicon International Inc., Temecula, Calif.) and quantified by using direct fluorescent antibody (DFA).
DNA controls were generated using genomic DNA isolated from specific serotypes belonging to the six different AdV species. Genomic DNA was amplified using species-specific primer pairs to yield PCR products for cloning. PCR products were cloned using the TOPO TA cloning kit for sequencing (Invitrogen, San Diego, Calif.). Briefly, PCR products were ligated into the pCR4-TOPO plasmid vector under standard conditions. Transformation of TOP10 competent cells with the plasmid and insert was carried out according to the manufacturer's instructions. Clones were verified for the specific PCR insert by restriction enzyme digestion as well as PCR-enzyme hybridization assay. DNA was quantified on a spectrophotometer to obtain the copy number and frozen at −70°C.
Species-specific oligonucleotide primers and probes were designed from conserved regions of the fiber gene of the six respective AdV species A to F available from GenBank (Table (Table1).1). The assay methodology was as previously described (9). Briefly, DNA was extracted from 400 μl of clinical specimen or DNA control using the Clontech Nucleospin tissue kit (BD Biosciences, Palo Alto, Calif.). Purified DNA was amplified by PCR using a Supermix containing the six pairs of primers and 2.5 U of Taq polymerase (Applied Biosystems, Foster City, Calif.). The PCR parameters were 95°C for 10 min, two cycles at 95°C for 1 min, 55°C for 30 s, and 72°C for 45 s, and 38 cycles at 94°C for 1 min, 60°C for 30 s, and 72°C for 30 s, and the reactions were then held at 72°C for 7 min. Following amplification, PCR products were purified using a Clontech Nucleospin extraction kit (BD-Biosciences). Hybridization solutions containing horseradish peroxidase-labeled probes corresponding to each AdV species (A to F) were added to 96-well streptavidin-coated microtiter plates, and the purified PCR products were then added to each well. These oligonucleotide probes represented highly conserved regions of the previously mentioned fiber gene. Following hybridization, plates were washed and tetramethyl benzidine substrate solution (Kirkegaard & Perry Laboratories, Gaithersburg, Md.) was added to each well. After 10 min, the reaction was stopped by adding 1 N H2SO4, and the optical density (OD) of each well was measured at 450 nm on a spectrophotometer. The positive cutoff value was calculated as four times higher than the negative control value and greater than or equal to an OD value of 0.400.
Serial dilutions of both recombinant DNA controls and whole virus were used to determine the assay's limit of detection (LOD). Ten-fold serial dilutions of DNA controls were made from 106 to 10 copies/ml in M4 (viral transport medium; Microtest, Snellville, Ga.) and tested. AdV stocks were diluted 10-fold from 103 to 10−3 TCID50/ml in M4 and tested with the Adenoplex.
AdV strains were diluted to 102 TCID50/ml in M4. All other viral strains, with the exception of influenza B virus, were diluted to 104 TCID50/ml. Influenza B virus and the bacterial strains tested were directly diluted 1:1,000 with M4. All dilutions were run through the Adenoplex to evaluate the specificity of the assay.
The specificity of the Adenoplex assay was tested using 23 AdV serotypes representing all six species, as well as other viruses and bacteria that cause similar respiratory symptoms (Table (Table2).2). Our assay was able to detect and specifically identify the species type for each of the AdV serotypes tested. Also, no cross-reactivity with other respiratory viruses or bacteria was observed.
The analytical sensitivity of the Adenoplex was tested using serial dilutions of recombinant DNA controls and whole virus. As shown in Table Table3,3, the LOD for the DNA controls AdV A, B, D, E, and F was 100 copies/ml and for AdV C it was 1,000 copies/ml. AdV species A, B, C, D, E, and F had LODs in terms of the TCID50/ml value equal to 10−2, 10−1, 10−1, 10−2, 10−1, and 10−2, respectively, when serial dilutions of virus stocks were tested.
Samples containing various combinations of AdV representative of the six different species were evaluated with the Adenoplex assay. Clinical specimens were spiked with either medium (1,000 TCID50/ml) and low (100 TCID50/ml) concentrations or equal (100 TCID50/ml) concentrations of the stock viruses. The results shown in Table Table44 indicate that dual AdV species infections were accurately detected in each sample tested. In specimens spiked with uneven concentrations of species D and E or of species E and F, slight inhibition of detection of one of the viruses was observed. The OD values for E and F samples that were spiked at the lower concentration (100 TCID50/ml) were slightly lower (2.8 and 2.0) than the >3.0 values for species D and E samples spiked at the higher concentration (1,000 TCID50/ml). This slight inhibition was only observed in dual infections, because when each of these AdV species was tested individually at either concentration (100 or 1,000 TCID50/ml) the OD values were >3.0 (data not shown). Furthermore, it is unlikely this slight inhibition would affect the detection of dual AdV infection in clinical specimens. No inhibition in detecting either species was observed when virus concentrations were equal.
We compared the Adenoplex assay to the diagnostic gold standard of cell culture (CPE) and DFA testing to evaluate our assay's ability to detect and distinguish between the different AdV species. Our multiplex PCR assay precisely detected AdV in all 120 spiked nasopharyngeal swab specimens and did not detect AdV in the 30 unspiked specimens, as shown in Table Table5.5. CPE alone was not sufficiently sensitive to detect AdV species A, B, and F (sensitivity, 75%). DFA performed on infected cells at day 14 detected 27 of the 30 CPE-negative samples, yielding a final culture sensitivity of 97.5%. In specimens spiked with group F AdV, 3 out of the 20 spiked specimens were positive by PCR analysis but negative by culture and DFA. Discrepant analysis on these samples was performed at another reference laboratory using a different PCR-based assay. All three samples were confirmed as true positives by the other laboratory's assay. Compared to culture or DFA and PCR, the Adenoplex specificity and sensitivity were 100%, with a 95% confidence interval of 97 to 100% (n = 150).
Human AdV are responsible for causing a variety of diseases affecting all organ systems. The 51 known AdV serotypes are grouped into six different species and, generally, the AdV serotypes within a species have similar tropisms, pathogenicity, latency, and occurrence (14). Species F AdV (serotypes 40 and 41) are the second leading cause of gastroenteritis in young children, and group A AdV (serotype 31) have also been implicated in gastroenteritis in infants. AdV types belonging to species C (serotypes 1, 2, and 5) cause acute respiratory disease in children, and types in groups B (serotype 7) and E (serotype 4) are leading causes of acute respiratory disease in military trainees. Highly contagious keratoconjunctivitis is primarily caused by AdV types belonging to species D (serotypes 8, 19, and 37). For immunocompromised individuals, especially bone marrow transplant recipients, species B (serotypes 11, 34, and 35) and C (serotypes 1, 2, and 5) AdV can cause severe and sometimes fatal disease. It is unknown how many different serotypes or species can infect someone at the same time during acute infection. We have observed dual infections (species A and C) in clinical specimens (unpublished data). Thus, an assay that is able to detect as well as differentiate between AdV species groups such as the Adenoplex can have important clinical relevance in diagnosis and for epidemiological purposes.
PCR methods of AdV detection offer a more sensitive and rapid alternative for diagnosis than traditional culture or immunofluorescence (1, 18, 20). Our data demonstrated that the Adenoplex assay was more sensitive (100%) and had equivalent specificity (100%) compared to viral culture. Three of the 20 clinical specimens that were spiked with species F AdV were positive by the Adenoplex but negative by culture and DFA. These three samples were spiked at the low concentration of 100 TCID50/ml. Discrepant analysis using a different PCR assay (with the hexon gene as the target) performed at another reference laboratory confirmed that the three specimens were indeed true positives. These results demonstrated that detection of AdV by PCR is more sensitive than culture and DFA. Furthermore, the Adenoplex achieved its sensitivity in 5 h, while culture and DFA took up to 14 days. Some clinical specimens may have even lower quantities of AdV (e.g., immunocompromised patients), and the Adenoplex's lower LOD compared to culture may result in larger sensitivity differences in clinical practice.
The Adenoplex assay has advantages over several of the PCR-based assays recently described in the literature. The Adenoplex primer pairs were designed to highly conserved areas of the fiber gene. Xu et al. found that when they used hexon gene primers from another recently described species-specific multiplex assay they had poor amplification of serotypes 11, 21, 34, and 35 of species B, yet their assay using fiber gene primers amplified all species B serotypes equally well (23). The Adenoplex accurately detected these four serotypes as well as all of the other serotypes tested. The fiber gene may have areas more highly conserved among AdV serotypes. However, additional clinical testing and sequencing would need to be done to determine this. The most effective diagnostic test for AdV is one that is easy and quick to perform. Several other PCR-based assays use restriction enzyme digestion (8, 14) and gel electrophoresis (19, 23) methods for detection. These techniques are time-consuming and limit the scope of detection. In contrast, the Adenoplex detection method of enzyme hybridization is rapid, provides increased sensitivity, and gives objective results (reported as OD readings). Amplified PCR products extracted from samples with low titers of whole organisms or recombinant controls give low, but definitely positive, OD values, whereas no bands are observed by gel electrophoresis (unpublished observations). Nested PCR assays have been shown to detect AdV with good sensitivity (1, 18) but involve a two-step PCR procedure increasing the time to complete the assay and chance of carryover contamination. The Adenoplex's sensitivity (at 100 and <1,000 copies of DNA/ml) is comparable to or better than that of the nested PCR assays, which detected between 400 and 2,500 copies/ml (1) and 640 copies/ml (18), respectively. In addition, the Adenoplex is a single-step PCR and can be completed in less time than a two-step nested PCR assay. Recent interest has been shown for real-time PCR assays, which have advantages over conventional PCR in terms of time and decreased risk of false-positive results due to contamination. However, the sensitivity of the Adenoplex was two times greater than that of a recently published real-time assay (13). Furthermore, the Adenoplex is able to differentiate between AdV species, whereas the real-time assay could only generically detect AdV.
Understanding the epidemiology and natural history of AdV infections in humans requires a rapid, sensitive, and specific diagnostic test. Proper management and treatment of AdV infections are equally dependent on effective methods of diagnosis. Same-day, rapid-turnaround PCR assays may provide cost-effective reporting that improves patient care and treatment. Recently it has been reported that detection of AdV in serum or blood may be a precursor of severe or disseminated disease. Timely and sensitive detection of AdV may serve as a basis for preemptive antiviral treatment (6, 16). Earlier AdV detection in pediatric transplant recipients by using PCR, compared with detection by culture, may result in reduced mortality (17). PCR may be the only means by which rapid and reliable diagnosis of AdV keratoconjunctivitis can be achieved, leading to a reduction in the overall cost of diagnosis (7). Finally, PCR is able to facilitate accurate identification of viruses that cause specific diseases. In the case of viral myocarditis, AdV, not enteroviruses as previously thought, were found to be the leading cause of disease (2).
In summary, we report here the development of a sensitive and specific multiplex PCR assay, the Adenoplex, for the detection of all human AdV serotypes as well as for identification of species types. This rapid and definitive method of diagnosis offers a powerful alternative approach to AdV detection and may improve clinical management, decrease hospitalization, and lessen medical costs for patients.
We thank Gerald Sedmak at the City of Milwaukee Health Department for his expertise and assistance in culturing the different AdV species and Pam Douglass, Tracy Enslow, Kristine Schraufnagel, and Katie Wilson for their technical assistance.