|Home | About | Journals | Submit | Contact Us | Français|
We evaluated the clinical applicability of a molecular serotyping method for determination of the cause of epidemic acute hemorrhagic conjunctivitis. Seventy conjunctival swab specimens from individuals involved in a nationwide acute hemorrhagic conjunctivitis outbreak were tested. Viral culture and a molecular biology-based assay were compared by directly using clinical specimens. On the one hand, virus culture was done to isolate the enteroviruses, and serotyping was done by a coxsackievirus A24 variant-specific PCR. On the other hand, the original clinical specimens were directly screened for enterovirus by reverse transcription (RT)-PCR with panenterovirus-specific primers. Enterovirus screening-positive specimens were subjected to RT-PCR for detection of the VP1 region of enterovirus, and the amplicons were sequenced. Molecular serotyping was done by calculating the pairwise identity scores for the sequences with the maximum identities to the sequences of known prototype enteroviruses. Thirty-two specimens (45.7%) were culture positive, whereas 37 specimens (52.8%) were screening PCR positive (P < 0.001). The VP1 regions were amplified from 21 of the 37 specimens (56.8%), and the products amplified from 9 specimens were appropriately sequenced. These nine sequences were homologous with the sequence of the coxsackievirus A24 variant. Molecular serotyping by direct use of clinical specimens without cell culture could be applied for the rapid identification of the causative agent of epidemic acute hemorrhagic conjunctivitis.
Human enteroviruses (EVs; family Picornaviridae) are classified into polioviruses, coxsackie A viruses, coxsackie B viruses, echoviruses, and EVs with specific consecutive numbers assigned to each virus on the basis of its pathogenesis in humans and experimental animals. There are 66 immunologically distinct serotypes (9). Infections with the EVs have a wide range of clinical outcomes, such as poliomyelitis, aseptic meningitis, hand-foot-mouth disease, herpangina, and acute hemorrhagic conjunctivitis (AHC). Because of these protean clinical manifestations and serotypes, the specific microbiologic detection of a virus is very important.
The conventional methods used for the isolation and identification of EVs are cell culture and the neutralization test with pooled antisera. These methods are, however, labor-intensive and time-consuming, and they require large pools of reference antisera. As a result, these methods can be performed in only limited laboratories. New strains of viruses are untypeable, and negative cell culture results for clinical specimens are nondiagnostic (9, 12). These disadvantages of conventional methods have resulted in the development of molecular biology-based methodologies as non-culture-based methods for virus detection (15, 17).
The VP1 region of the EV gene contains a major neutralization epitope, and its sequence is known to correlate well with the classical serotypes (13). A recent study showed that sequencing of the VP1 region by using primers containing deoxyinosine and mixed bases for codon degeneracy could identify almost all serotypes, and this new method was almost 100% specific (13). This molecular serotyping method is such a promising diagnostic tool that it could replace the laborious conventional cell culture method (8, 10, 11, 12, 13), despite its own limitations.
Previous studies of molecular serotyping, however, have used viral stocks from cell cultures, and little is known about how this molecular serotyping method could replace the multistep cell culture-based method. To evaluate the clinical applicability of the molecular serotyping method, we tried to identify the causative agent of an outbreak of AHC by the molecular serotyping method by directly using clinical specimens.
This brief study was done as a part of our investigation of a nationwide AHC outbreak. The specimens used in this study consisted of 70 conjunctival swab specimens from patients involved in one AHC outbreak caused by the coxsackievirus A24 variant (CA24v), as already suggested in our previous study (14). Viral culture and the molecular biology-based assay were compared by performing the assays directly with clinical specimens. On the one hand, virus culture was done to isolate the EVs, and serotyping was done by a CA24v-specific PCR. On the other hand, the original conjunctival swab specimens were directly subjected to PCR and amplicon sequencing for molecular serotyping. All the specimens were screened for adenovirus by a previously described PCR method (2) to evaluate whether the specimens had concurrent infections. Six specimens (8.6%; four specimens infected with adenovirus only and two specimens coinfected with adenovirus and EV) were positive by an adenovirus-specific nested PCR.
Conjunctival swab specimens were taken from 70 AHC patients at a community-based eye clinic in Seoul, South Korea, from 3 to 16 September 2002, when a nationwide large epidemic of AHC was at its peak. The inferior tarsal conjunctiva and fornix were swabbed with a Dacron swab soaked with 2 ml of Eagle's viral transport medium. All conjunctival swab specimens were vortexed, and the swab sticks were removed. Part of each swab specimen was subjected to cell culture within 4 h, and then the remaining sample was frozen at −70°C. The frozen specimens were later thawed and were used within 4 months.
Conjunctival swab specimens were inoculated onto two cell lines (HeLa cells and MRC-5 cells), and the cells were observed for the development of cytopathic effects. The cultures were held for 2 weeks. Serotyping of the virus isolates was carried out with the culture supernatants by a CA24v-specific PCR (16). Although this is not the classical approach, it was thought to be a reasonable one, because the etiologic agent and its homogeneity have been suggested previously (14).
As a counterpart to virus culture, all original conjunctival swab specimens were directly screened by reverse transcription (RT)-PCR with pan-EV-specific primers that anneal to highly conserved sequences of the 5′ untranslated region of the EV genome to select EV-positive specimens (12, 18). The final product was 114 bp. Then, EV screening-positive specimens were subjected to RT-PCR with primer 188 (forward primer; 5′-ACI GCI GTI GAR ACI GGN G) and primer 222 (reverse primer; 5′-CIC CIG GIG GIA YRW ACA T) to amplify the 339-bp VP1 region of EV for sequencing (12). “I” denotes inosine. IUB codes are as follows: Y is T or C; W is A or T; and R is G or A.
The latter RT-PCR was done as described below. RNA was extracted from 140 μl of each swab specimen by using a QIAamp viral RNA isolation kit (Qiagen Inc., Valencia, Calif.), and the RNA was then eluted in 60 μl of elution buffer. RT was done in a reaction volume of 20 μl containing 1× RT buffer, 1 mM deoxynucleoside triphosphates, 10 mM dithiothreitol, 1× random hexamer, 50 U of Moloney murine leukemia virus reverse transcriptase (Roche Diagnostics GmbH, Mannheim, Germany), 40 U of human placental RNase inhibitor (Promega, Madison, Wis.), and 5 μl of RNA template. The reaction was initiated at 30°C for 10 min, incubated at 42°C for 50 min, and inactivated at 95°C for 2 min. PCR was done with a reaction volume of 50 μl containing 1× reaction buffer, 0.2 mM deoxynucleoside triphosphates, 5 μl of dimethyl sulfoxide, 25 pmol of each primer (primers 188 and 222), 2.5 U of Taq polymerase (Roche Diagnostics GmbH), and 4 μl of template cDNA. The reaction was initiated at 94°C for 5 min; and thermocycling (GeneAmp PCR system 2400; Perkin-Elmer, Norwalk, Conn.) was performed for 40 cycles of 94°C for 45 s for denaturation, 44°C for 45 s for annealing, and 72°C for 45 s for extension, with a final extension at 72°C for 7 min.
The final 339-bp PCR products were purified with a QIAamp DNA purification kit (Qiagen Inc.), and the sequencing reactions were performed in an MJ Research gradient cycler by using a Dye Terminator Cycle Sequencing Ready Reaction kit with AmpliTaq DNA polymerase (Applied Biosystems, Foster City, Calif.), according to the protocol of the manufacturer. Sequences that matched the VP1 sequences of the test strains were identified by using the BLAST program (www.ncbi.nlm.gov/BLAST/) from GenBank. The percent identity scores for the VP1 sequences of the test strains with the EV sequences in GenBank that showed the maximum identity, as well as the next closest identity, were calculated by using the ClustalW program. If the pairwise identity scores for the VP1 sequences of the test strains were 75% or more with respect to any particular EV sequence in GenBank, then it was identified as a homologous serotype (11).
The sequences from this study are available in GenBank with the following accession numbers: AY296248 to AY296251 and AF545847 to AF545848.
Table Table11 compares the results of viral culture and those of the EV-specific PCR tests performed directly with the clinical specimens. Of the 70 conjunctival swab specimens tested, 32 (45.7%) were culture positive, with extensive cytopathic effects detected mostly 2 to 5 days after inoculation, and all the isolates cultured were typed as CA24v by PCR. Of the 32 specimens that were culture positive, 31 (97%) had positive amplifications by the EV screening PCR with pan-EV-specific primers; 1 sample (3%) was PCR negative. Of the 38 specimens that were culture negative, 6 were EV screening PCR positive and 32 were negative. Of the 37 EV screening PCR-positive specimens, 6 (16%) did not yield EV by viral culture. Detection of EV by screening PCR resulted in a significant increase (16%; P < 0.001) in the number of positive specimens compared to the number found to be positive by viral culture. Of the 37 EV-screening PCR positive specimens, 21 (56.8%) were positive by RT-PCR amplification of the VP1 region with primers 188 and 222.
Twenty-one amplicons of the VP1 region were subjected to nucleotide sequencing, and 9 of these yielded appropriate sequences. The other 12 products could not be sequenced due to inadequately low concentrations of the amplified products. The GenBank sequence with the maximum identity with the test sequences was that of CA24v, and the pairwise identity scores were all 85%, confirming that CA24v is the serotype whose sequence is homologous with the test EV sequences. No other GenBank sequence had a pairwise identity score greater than 75%.
The molecular serotyping method (6) could be a good alternative to the “gold standard” cell culture-neutralization method in view of its easy accessibility and expected sensitivity and the good correlation of the results with those obtained by the classical method. Previous studies of molecular serotyping were done to evaluate the correlation between this molecular biology-based method and the cell culture-neutralization method and to identify serologically untypeable strains by the molecular biology-based method (8, 10, 11, 12, 13). The substrates for this molecular biology-based technique were viral isolates that had been prepared from clinical specimens by the conventional cell culture method. When the high degree of specificity of the molecular serotyping method for determination of the EV serotype is considered, the high degree of sensitivity could yield a diagnostic result for a clinical specimen within a few days, and this could make the new method a good diagnostic tool in this era of easy accessibility to molecular biology-based methods.
In this study, screening of the conjunctival swab specimens with pan-EV-specific primers yielded positive amplifications for 52.8% (37 of 70) of the specimens, whereas cell culture provided positive results for 45.7% (32 of 70) of the specimens. Given that prior studies of EV conjunctivitis by cell culture showed about 50% positive results with conjunctival swab specimens (1, 4, 5, 6, 7), we thought that the initial quality of the specimens was appropriate for their use in this study. RT-PCR for the VP1 region of the 37 EV screening PCR-positive specimens yielded products for 21 of the specimens, which is a rate 56.8% (21 of 37 specimens) and which is a sensitivity about half that of the screening method. The optical densities of these 21 amplification products varied widely. The products from only 9 of the 21 specimens could be sequenced. The concentrations of the nucleotides amplified from the remaining 12 specimens were too low for sequencing. Of the 31 culture- and screening PCR-positive specimens, 11 were negative for amplification of the VP1 region. The failure of PCR to detect the culture-positive samples may be explained by storage conditions, laboratory manipulations that resulted in degradation of the EV RNA, and/or the presence of amplification inhibitors. Viral culture with the original fresh specimens was done without delay or freezing of the specimens. The remaining parts of the specimens, however, were frozen and the specimens were subjected to several thawing-freezing processes. After the initial EV screening RT-PCR, there was some time lag until the following PCR tests and molecular serotyping tests were performed. We adjusted the reaction temperature and the time for the optimal output, and the test conditions used in this study provided the best results. The EV screening PCR and the VP1 region-specific PCR were repeated with two of the specimens known to have strongly positive results by both PCR tests by using 10-fold serial dilutions of the specimens. This resulted in marginal or negative amplifications with dilutions of 10−4 to 10−5. These considerations suggested that degradation of the EV RNA was the most possible explanation for the inadequate amplification of the VP1 region and the subsequent failure of amplicon sequencing. Therefore, care should be taken to minimize the time lag time during the handling of RNA specimens and the frequency of specimen manipulation.
A recent study (3) that used the molecular serotyping method for virus detection during an AHC outbreak provided good results. Of 26 conjunctival swab specimens tested, 20 (77%) were positive for EV by screening assays; of these, 19 (95%) were identified as CA24 by sequencing of the VP1 region. They used a one-step RT-PCR (3); otherwise, they used a methodology similar to ours.
Our study showed that the causative agent of an AHC outbreak was EV by the EV screening PCR, which was superior to viral culture. Although molecular serotyping results were available for some of the specimens (9 of 37; 24%), we could conclude that CA24v was the causative serotype of EV because only that agent was identified and no positive results for any alternative agent were obtained. Our study has no implications about the application of molecular serotyping method to the detection of infections caused by other EVs, but we think that this method deserves full consideration.
We thank Young-Ae Yoo (Hamchun Eye Clinic, Seoul) and Sung-Han Kim, Cheol-In Kang, Wan-Beom Park, and Bo-Bin Lee (Department of Internal Medicine, Seoul National University College of Medicine, Seoul) for help during sample collection and processing.