|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Molecular Diagnostics and Genetics |
1 Institute of Virology, Helmut-Ruska-Haus, Charité Campus Mitte, Berlin, Germany.
2 Robert Koch-Institut, Zentrum für Biologische Sicherheit 1, Berlin, Germany.
3 Institute of Virology, Slovak Academy of Sciences, Dubravska cesta 9, Bratislava, Slovakia.
aAddress correspondence to this author at: Robert Koch-Institut, Nordufer 20, 13353 Berlin, Germany. Fax 49-(0)30-18754-2605; e-mail nitschea{at}rki.de.
| Abstract |
|---|
|
|
|---|
Methods: We developed 5 real-time RT-PCR assays, 3 of which are specific for the individual European hantaviruses Dobrava, Puumala, or Tula virus. Two additional assays detect the Asian species Hantaan virus together with Seoul virus and the American species Andes virus together with Sin Nombre virus. Pyrosequencing was established to provide characteristic sequence information of the amplified hantavirus for confirmation of the RT-PCR results or for a more detailed virus typing.
Results: The real-time RT-PCR assays were specific for the respective hantavirus species and optimized to run on 2 different platforms, the LightCycler and the ABI 7900/7500. Each assay showed a detection limit of 10 copies of a plasmid containing the RT-PCR target region, and pyrosequencing was possible with 10 to 100 copies per reaction. With this assay, viral genome could be detected in 16 of 552 (2.5%) specimens of suspected hantavirus infections of humans and mice.
Conclusions: The new assays detect, differentiate, and quantify hantaviruses in clinical specimens from humans and from their natural hosts and may be useful for in vitro studies of hantaviruses.
| Introduction |
|---|
|
|
|---|
Infection of humans with hantaviruses induces 2 severe diseases: hemorrhagic fever with renal syndrome (HFRS)1 and hantavirus cardiopulmonary syndrome (HCPS). The Old World hantaviruses Dobrava (DOBV), Puumala (PUUV), and Tula (TULV) cause HFRS in Europe (6)(7)(8), whereas Hantaan virus (HNTV) and Seoul virus (SEOV) are the most prominent HFRS pathogens in Asia (9). The New World hantaviruses Andes (ANDV) and Sin Nombre (SNV) cause HCPS in the Americas (10)(11)(12).
Infection with PUUV, the Central European DOBV lineage DOBV-Aa, and TULV varies from moderate disease to subclinical courses, and infection with the hantavirus DOBV lineage from the Balkans, DOBV-Af, or the hantavirus species SEOV, HNTV, ANDV, and SNV affects various organs and can lead to more severe and even fatal disease (6)(11)(13)(14).
Standard diagnostic methods for hantaviruses are based on serologic and immunologic techniques [for a review, see Refs. (1), (15), (16)]. These serologic assays do not take into account the viremic status of the patient, and serologic cross-reactivity between different hantaviruses hinders species differentiation. Other serotyping approaches, such as the characterization of neutralizing antibodies by focus-reduction neutralization test (17), are labor intensive, time-consuming, and require BSL3/BSL4 safety conditions.
Consequently, a specific and sensitive diagnostic method for the detection, differentiation, and quantification of hantaviruses is required to detect hantavirus in an early state of infection, before the onset of clinical symptoms. Today, real-time PCR is considered the gold standard for rapid pathogen detection (18), and several assays were recently reported to detect individual hantavirus species such as PUUV (19) and DOBV (20). A recently published real-time reverse transcription (RT)-PCR assay based on degenerated primers amplified RNA of DOBV, PUUV, HNTV, and SEOV but differentiated these hantaviruses by species-specific 5' nuclease probes (21).
Based on real-time RT-PCR, this study describes the development of 3 real-time RT-PCR assays for the specific detection of the European hantaviruses DOBV, PUUV, and TULV. In addition, 2 real-time RT-PCR assays were established for the simultaneous detection of the Asian species HNTV and SEOV or the American species ANDV and SNV in 1 reaction tube each. For each real-time RT-PCR, a pyrosequencing assay was established to enable virus typing.
Pyrosequencing is a new technique in which the enzymatic incorporation of different nucleotides that are added separately and are complementary to the sequenced DNA strand is monitored online. For each incorporated nucleotide, a light signal is generated and presented as a peak histogram, called a pyrogram. Because of the loss of signal intensity with an increased number of incorporated nucleotides, pyrosequencing has its strength in the detection of sequences of up to 80 bases and the identification of single nucleotide polymorphisms. Usually this maximal length is sufficient for the identification and typing of pathogens, as shown for the highly diverse group of papillomaviruses (22). Recently, pyrosequencing was also used for determination of HIV resistance mutations (23) or influenza virus resistance to adamantane (24). In this study, we applied pyrosequencing to generate short sequence stretches of the respective real-time RT-PCR products, allowing hantavirus species differentiation.
| Materials and Methods |
|---|
|
|
|---|
pcr templates: isolation of viral rna, reverse transcription, and plasmid cloning
We performed real-time RT-PCR with the hantaviruses DOBV [DOBV-Aa, strain Slovakia (DOBV/SK) and DOBV-Af, strain Slovenia (DOBV/Slo)], PUUV (strain Vranica-Hällnäs), TULV (strain Moravia), HNTV (strain 76–118), and SEOV (strain 80–39). Because propagation of ANDV and SNV was not possible due to BSL4 restrictions, we used yeast cells producing recombinant nucleocapsid proteins of ANDV (strain AH-1) (26) and SNV (strain 3H226) (27) to produce virus-specific RT-PCR templates. All other hantaviruses were propagated in Vero E6 cells, and the viral titers were determined by focus assay as described (17).
Total virus RNA was isolated from 140 µL virus-containing cell culture supernatant, using the QIAamp Viral Mini Kit according to the manufacturers instructions (Qiagen). cDNA was produced by reverse transcription in a total volume of 20 µL containing 4 µL 10x buffer, 0.1 µL 0.1 mol/L dithiothreitol, 3 µL 2.5 mmol/L deoxynucleotide triphosphate, 5.0 pmol random hexamers, 0.5 µL 40 units/µL RNase inhibitor, 0.5 µL 200 units/µL Moloney murine leukemia virus reverse transcriptase, and 10 µL purified RNA. Samples were incubated at 25 °C for 10 min, 42 °C for 10 min, and 96 °C for 6 min.
For calibration of all assays, we cloned the respective PCR amplicons into a TOPO TA cloning vector according to the manufacturers instructions (Invitrogen). After plasmid preparation, we determined the DNA concentration by use of a spectrophotometer at 260 nm and calculated the corresponding copy number. Calibration curves were generated by amplification of 10-fold serial dilutions of 105 to 101 plasmid copies per reaction in a constant background of 1 ng/µL
DNA or mouse DNA.
real-time rt-pcr
Degenerated oligonucleotide primers and the minor groove binder 5' nuclease probes were designed to bind to a highly conserved region within the S-segment of the hantavirus genome. The sequences, genome locations, and annealing temperatures of oligonucleotide primers and minor groove binder probes are listed in Supplemental Data Table 1, which accompanies the online version of this article at http://www.clinchem.org/content/vol53/issue11 .
Real-time RT-PCR conditions for 1-step and 2-step RT-PCR were established for the ABI 7900/7500 series (Applied Biosystems) and the LightCycler (Roche Applied Science). All reaction conditions are listed in detail in Supplemental Data Table 2.
specificity
In addition to nucleic acids of various hantavirus species, the specificity of the assays developed was evaluated with nucleic acids of several other pathogens, including influenza A and B virus, yellow fever virus, dengue fever virus, tick-borne encephalitis virus, rift valley fever virus, Marburg virus, Ebola virus, adenovirus, poxviruses, and Bacillus anthracis, as well as cDNA of mice and humans (Table 1
). All of these nucleic acids had previously tested positive with specific assays (data not shown), indicating target concentrations of >105 per reaction.
|
detection limit
We determined the detection limits and imprecision of each assay by use of 2-step RT-PCR, measuring 3 different concentrations of viral cDNA, reflecting high, medium, and low viral load, and in addition serial dilutions of plasmids (105 to 101), each measured in quadruplicate on the TaqMan 7900 and the LightCycler. All experiments were repeated on 3 consecutive days to determine interassay variability. Linearity and calibration curves were calculated with Microsoft Excel 2003.
To determine the sensitivity of pyrosequencing, 10-fold serial dilutions of the corresponding plasmids were amplified by real-time PCR and directly subjected to pyrosequencing.
pyrosequencing
We performed pyrosequencing analysis by use of a PyroMark ID System (Biotage) directly after real-time RT-PCR was finished. Real-time RT-PCR was performed with 1 biotinylated primer per assay to allow efficient single-strand DNA preparation by using streptavidin-coated Sepharose beads. The single-stranded amplicon was annealed to a specific sequencing primer that is usually different from the amplification primers. For all pyrosequencing assays, the biotinylated real-time RT-PCR primers and sequencing primers are given in Supplemental Data Table 1. In the case of ANDV, a non-extendable "blocking primer" had to be designed and included to the sequencing mix to inhibit 3'-end hairpin formation of the target by unwanted sequence generation (28). Pyrosequencing reactions were performed using Pyrogold SQA reagents (Biotage) with the complete real-time RT-PCR product according to the manufacturers instructions.
| Results |
|---|
|
|
|---|
Moreover, additional specificity tests performed with further pathogens and mouse and human nucleic acids showed no cross-reaction to these targets (Table 2
).
|
performance characteristics
The PCR efficiency for each assay is given in Table 3
. All uniplex assays displayed a dynamic detection range from 105 to 101 target copies, with a correlation between r2 = 0.98 and r2 = 0.99 when performed as 2-step real-time RT-PCR for both platforms and a detection limit of
10 copies of a plasmid containing the RT-PCR target region (see Table 3
). Representative amplification curves are shown in Supplemental Data Fig. 1.
|
For the determination of the reproducibility (intra- and interassay variability), cDNA from clinical specimens with CT values between 29 and 33 were measured in quadruplicate on 3 consecutive days. For all assays (except for the HNTV/SEOV interassay variation with HNTV cDNA as template) the range of the threshold cycle (CT) values was <1 CT. The combination of the uniplex PUUV and the uniplex DOBV assay into a multiplex reaction caused no adverse effects on the reproducibility (data not shown). Table 3
summarizes typical results of the variability and efficiency testing.
To shorten the hands-on time, all reactions were also established as 1-step real-time RT-PCR. For the respective reaction conditions, see Supplemental Data Table 2.
pyrosequencing
To confirm the real-time RT-PCR results of hantavirus species–specific assays or to identify the hantavirus species in specimens that came up positive in the multiplex assays (for the detection of DOBV/PUUV, HNTV/SEOV, or ANDV/SNV), we applied the pyrosequencing technique subsequently to real-time RT-PCR. Using the pyrosequencing primers (see Supplemental Data Table 1), virus-specific sequences between 24 and 67 bases for respective hantavirus species could be generated in <1 h after the real-time RT-PCR run. Sequence comparison allowed the explicit classification of the virus species amplified by the 3 single (DOBV, PUUV, and TULV) or the 3 multiplex (DOBV/PUUV, HNTV/SEOV, and ANDV/SNV) real-time RT-PCR assays. Fig. 1
shows a representative pyrogram of the HNTV/SEOV assay performed with cDNA of HNTV or SEOV, respectively.
|
We determined the sensitivity of the additional pyrosequencing step by analyzing 10-fold dilutions of the corresponding plasmids by pyrosequencing directly after amplification by real-time PCR. The best sensitivity was obtained for PUUV and SEOV, with a sequencing detection limit of 10 amplified plasmid copies per reaction. For all other viruses, sequences analyzable by comparison to sequence databases could be obtained when
100 copies of the corresponding plasmid were amplified and subjected to pyrosequencing.
clinical specimens
Finally, 228 clinical specimens from humans suspected to be hantavirus positive and 324 mice specimens were tested by nested RT-PCR and the new real-time RT-PCR assays in duplicate. The results are summarized in Table 2
. In total, 7 of 228 (3.1%) of the human specimens tested positive in nested and real-time RT-PCR. Of the mice specimens, 10 of 324 (3.1%) tested positive in nested RT-PCR and 9 of 324 (2.8%) in real-time RT-PCR assays. Two human acute sera of HFRS patients from Greece, kindly provided by Anna Papa (Thessaloniki, Greece), were detected to be positive for DOBV, and 5 human acute sera of HFRS patients and 5 bank vole (Clethrionomys glareolus) specimens from Germany were positive for PUUV. Of the mice specimens from Slovakia, 2 striped field mice (Apodemus agrarius) and 1 yellow-necked mouse (Apodemus flavicollis) were positive for DOBV, and 1 common vole (Microtus arvalis) was positive for TULV. Quantification in human serum revealed virus loads of 320 to 5200 genome equivalents/mL, and in mouse serum, 1000 to 107 genome equivalents/mL. All real-time RT-PCR results could be confirmed by pyrosequencing yielding analyzable sequences of up to 30 bases. Although nested PCR assays are usually expected to have the best sensitivity, our data indicate nearly the same sensitivity for the newly developed real-time RT-PCR assays. Clinical specimens infected with other hantaviruses, like ANDV or SNV, were not available.
| Discussion |
|---|
|
|
|---|
In contrast to the recently published hantavirus-specific RT-PCR assays, our assay is suitable for a fast screening of the major hantaviruses in Europe, Asia, and the Americas and enables rapid and exact virus typing in clinical specimens. Three specific single real-time RT-PCR assays were developed for the European hantaviruses DOBV, PUUV, and TULV and 2 real-time RT-PCR assays for the detection of (a) the Asian hantaviruses HNTV and SEOV and (b) the American hantaviruses ANDV and SNV in 1 reaction tube each.
All assays featured the major benefits of real-time PCR–speed, a very low detection limit of 10 genome equivalents per reaction, and a linear quantification range from 105 to 101. Despite the pronounced diversity of the RNA genome of hantaviruses that resulted in the design of highly degenerated primers, the detection limits and the assay imprecision were nearly identical for all assays under optimized reaction conditions. In addition, the uniplex assays for the European hantaviruses PUUV and DOBV could be combined in a multiplex RT-PCR with no significant loss in sensitivity and specificity, making it a feasible tool for the rapid screening of clinical specimens in Europe. To reduce specimen handling, all real-time RT-PCR assays could be performed as 1-step RT-PCR with common reagents, resulting in comparable specificity, sensitivity, and reproducibility.
For hantavirus typing or confirmation of the real-time RT-PCR results, pyrosequencing could be performed directly after RT-PCR amplification, including the differentiation of HNTV from SEOV as well as ANDV from SNV, which is not possible by real-time RT-PCR only.
To demonstrate the usefulness of the new developed assays, we tested 228 human clinical specimens and 324 mouse specimens. Altogether 16 of 552 (2.5%) of the specimens were positive for hantavirus RNA. Two human sera specimens were positive for DOBV, 5 human sera specimens as well as 5 mouse specimens were positive for PUUV, 3 mouse specimens were positive for DOBV, and 1 mouse specimen was positive for TULV. As expected, no clinical specimens were positive for HNTV/SEOV or ANDV/SNV.
Due to the fact that species typing in hantavirus infection is not possible during the early stage of infection by serologic methods, these assays offer a detection limit that promises the identification of the respective hantavirus directly after onset of clinical symptoms. However, it should be mentioned that hantavirus RNaemia is usually short-lasting and even undetectable in certain patients (1), which is reflected by the small number of PCR-positive specimens. Therefore the real-time assays for the major hantaviruses from Europe, Asia, and America should be understood as a supplemental technique to serologic diagnostic procedures for hantaviruses.
The real-time RT-PCR assays developed are specific for the desired hantavirus species even in concentrations close to the detection limit of
10 copies per reaction. By combining the specific assays for the clinically most relevant hantaviruses in Europe, DOBV and PUUV (31)(32), an expeditious screening of clinical specimens is possible. If necessary for confirmation of the real-time RT-PCR results or for differentiation of ANDV and SNV, as well as HNTV and SEOV, rapid pyrosequencing can be applied.
| Acknowledgments |
|---|
Financial disclosures: None declared.
Acknowledgments: We are grateful to Heike Lerch, Brita Auste, and Julia Tesch for excellent technical assistance and to Ursula Erikli for critically reading the manuscript.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
J. A. Jordan, J. Jones-Laughner, and M. B. Durso Utility of Pyrosequencing in Identifying Bacteria Directly from Positive Blood Culture Bottles J. Clin. Microbiol., February 1, 2009; 47(2): 368 - 372. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |