ABSTRACT
Environmental virus surveillance was conducted at two independent sewage plants from urban and rural areas in the northern prefecture of the Kyushu district, Japan, to trace polioviruses (PVs) within communities. Consequently, 83 PVs were isolated over a 34-month period from April 2010 to January 2013. The frequency of PV isolation at the urban plant was 1.5 times higher than that at the rural plant. Molecular sequence analysis of the viral VP1 gene identified all three serotypes among the PV isolates, with the most prevalent serotype being type 2 (46%). Nearly all poliovirus isolates exhibited more than one nucleotide mutation from the Sabin vaccine strains. During this study, inactivated poliovirus vaccine (IPV) was introduced for routine immunization on 1 September 2012, replacing the live oral poliovirus vaccine (OPV). Interestingly, the frequency of PV isolation from sewage waters declined before OPV cessation at both sites. Our study highlights the importance of environmental surveillance for the detection of the excretion of PVs from an OPV-immunized population in a highly sensitive manner, during the OPV-to-IPV transition period.
INTRODUCTION
Poliovirus (PV) is a nonenveloped, positive-sense, single-stranded RNA virus belonging to the genus Enterovirus of the family Picornaviridae (1, 2). PV possesses a relatively small icosahedral particle structure (∼30 nm in diameter) composed of four different capsid proteins, including VP1, where most antigenic epitopes are located (3, 4). Similar to other non-polio enteroviruses (NPEVs), PV is transmitted via the fecal-oral route and efficiently replicates in the intestinal tract (3). During PV infection, the virus is excreted from the human gut into the stool for ∼2 months (5–7). Although most PV infections are asymptomatic, patients can develop poliomyelitis following viremia in some cases, resulting in residual paralysis (8).
Since the live oral poliovirus vaccine (OPV) was introduced in many industrial countries in the 1960s, polio epidemics have been successfully controlled. In 1988, the World Health Assembly resolved to eradicate polio by launching the Global Polio Eradication Initiative (GPEI). Large-scale OPV immunization resulted in a drastic reduction in the number of poliomyelitis cases. To date, the only countries where polio is endemic are Nigeria, Pakistan, and Afghanistan (http://www.polioeradication.org/Dataandmonitoring.aspx), and the WHO is closely monitoring neighboring countries at an increased risk of reemergence of wild PV or vaccine-derived poliovirus (VDPV) to maintain a polio-free situation. In Japan, the last indigenous wild PV was isolated from a single patient with poliomyelitis in 1980 (9). In an effort to keep Japan polio free, OPV has been used for routine immunization for the last 50 years in Japan. It was scheduled twice for children between 3 and 18 months of age at an interval of >6 weeks, and children were immunized mainly in the spring and autumn seasons at schools and hospitals. To minimize the risk of vaccine-associated paralytic poliomyelitis (VAPP) due to OPV, a standalone conventional inactivated poliovirus vaccine (IPV) was introduced in September 2012. Thereafter, a Sabin-derived IPV in combination with the diphtheria-tetanus-pertussis vaccine (DTP-sIPV) (10) was introduced in November 2012 for routine immunization (three doses administered to 3- to 12-month-old babies, with one booster dose between the ages of 12 and 15 months, after the third immunization). Alarmingly, OPV national coverage declined to 67.2% for the immunization period of spring 2012, prior to the transition to IPV (http://www.mhlw.go.jp/bunya/kenkou/polio/). The refusal of OPV immunization was most likely due to the growing public concern about VAPP (11, 12). Consequently, the risk of PV infection has increased among unvaccinated children and in the larger population.
Environmental surveillance is a highly sensitive method for detecting enteroviruses such as PVs in environmental samples, and this practice has been adopted by many countries and regions worldwide (13–19). It is critically important to routinely monitor sentinel sites for the emergence of novel VDPV strains and the importation of wild PV from countries where the disease is endemic. In the global effort to eradicate polio, IPV immunization will be introduced before trivalent OPV cessation to minimize the risk of spread of VAPP to susceptible individuals within the population (20). We recently designed a comprehensive monitoring system for the surveillance of enteric viruses at sentinel hospitals (21) and for the determination of the relationship between environmental and patient surveillance. This study led to the isolation of enteroviruses and PVs. Here we report the prevalence of PVs in sewage water from two locations in Japan during the OPV-to-IPV transition period. Our results provide valuable information, at the local community level, on the impact of the transition period of PV immunization, with considerations on how OPV can be safely discontinued at the global level.
MATERIALS AND METHODS
Sample collection.Influent wastewater was obtained from two sewage disposal plants (plants T and Y) located in the northern area of Kyushu, Japan. The sanitation coverage in the area, defined as the percentage of the population connected to the sewage system relative to the entire population, is 61% (http://www.qsr.mlit.go.jp/n-park/city/index_e02_a.html). Plant T is located in an urban area and has high sanitation coverage (≥90%). Plant Y is located in a rural area and has low sanitation coverage (≤10%). The watershed populations of the two areas were similar. Approximately 190,000 persons live around plant T, and ∼180,000 persons live around plant Y. The child populations of the recommended age for OPV immunization (0 to 2 years old) were ∼2,300 children in the plant T watershed population and 1,400 children in the plant Y watershed population (Table 1). Every first week of the month, 1,000 ml of wastewater was routinely obtained at each plant and stored at 4°C until further processing.
Comparison of the appearance and isolated serotype of polioviruses from two sewage plants
Treatment of wastewater samples.The samples were concentrated as described previously (14, 22, 23). In brief, the collected wastewater was centrifuged at 1,500 × g (30 min at 4°C). Following this, MgCl2 was added to the resulting supernatant to a final concentration of 0.05 M, and the solution was adjusted to pH 3.5 with HCl. The supernatant was then applied to a negatively charged mixed cellulose ester filter (0.45-μm-pore-size filter; Advantec, Tokyo, Japan). The filter holder was equipped with a PST-1000 digital tube pump (Iwaki, Tokyo, Japan). The membrane was ground in 10 ml of 3% beef extract with a MediFASTH2 homogenizer (Omni International) to elute the bound viruses. After membrane grinding, the eluent was centrifuged at 16,000 × g for 30 min to remove debris and filtered with a 0.22- or 0.45-μm-pore-size polyvinylidene difluoride (PVDF) membrane filter (Millipore). The final concentrates were stored at −20°C until further analysis.
Virus isolation.The procedure for virus isolation involved two cultivation steps. First, 100 μl of each concentrate was inoculated into four wells of five cultured cell lines (Vero-E6, LLC-MK2, HEp-2, FL, and RD-18S) grown and maintained in 24-well plates containing Dulbecco's modified Eagle's medium (MEM; Sigma-Aldrich). Three cell lines (LLC-MK2, HEp-2, and FL) were purchased from Dainippon Pharmaceutical (Japan). Each cell line has a different range of sensitivity for virus isolation, and the RD-18S cell line was previously described to be advantageous for the isolation of coxsackievirus A (24). The cell lines were observed for 7 days, and samples with no cytopathic effect (CPE) were passed onto a new well plate and observed for another 7 days. During the observation period, the viruses isolated from the cell culture showed a CPE. Second, the supernatant of each CPE-positive culture fluid sample (100 μl) was inoculated onto L20B cells maintained in MEM for selectively isolating PVs and further studying CPE for 7 days (25, 26). For the screening test for non-polio enteroviruses (NPEVs), reverse transcription (RT), conventional PCR, and direct sequencing were carried out on all CPE-positive agents, as previously described (27).
Serotype identification and VP1 sequencing of PVs.For CPE-positive culture fluid samples after L20B cell inoculation, we performed a neutralization test (NT) with poliovirus type-specific antisera (Denka Seiken, Tokyo, Japan), in order to isolate each serotype from the poliovirus mixtures (14).
Total viral RNA was extracted from poliovirus isolates by using the QIAcube automated DNA/RNA purification system (Qiagen, Tokyo, Japan) on the basis of the QIAamp viral RNA minikit procedure (Qiagen), according to the manufacturer's instructions. RT-PCR was performed on the VP1 region of the viral genome by using a primer set specific for PV, UG1/UC11 (28), and the One Step RT-PCR kit (Qiagen). Direct sequencing was performed on samples with effective amplification confirmed by a single and specific gel band by electrophoresis. The PCR amplicons were enzymatically purified with Illustra ExoStar (GE Healthcare, Tokyo, Japan). The sequencing reactions were conducted by using the BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems, Tokyo, Japan). The sequenced products were purified by using the BigDye XTerminator purification kit (Applied Biosystems) and sequenced with the 3130xl genetic analyzer (Applied Biosystems). The full-length VP1 sequence of PV was aligned by using MEGA5 software (29). The VP1 sequences of the reference vaccine strains used in this study were those of Sabin 1 (S1) (GenBank accession number AY082688; 906 bp), Sabin 2 (S2) (accession number AY082679; 903 bp), and Sabin 3 (S3) (accession number AY082683; 900 bp).
Nucleotide sequence accession numbers.The entire VP1 nucleotide sequences of the 83 PV isolates in this study were deposited in the GenBank database under accession numbers AB829540 to AB829549, AB829551, AB829553 to AB829562, AB829564 to AB829572, AB829574 to AB829600, AB921169 to AB921180, and AB980981 to AB980994.
RESULTS
Isolation of PVs from sewage water.Concentrates from the wastewater collected at the two sewage plants in the northern prefecture of the Kyushu district, Japan, were inoculated onto five different cell lines. During the 34-month sampling period, 446 CPE-positive cultures were identified by using these 5 cell lines. After further reinoculation of all CPE-positive samples into L20B cultures, we confirmed CPE in 103 cultures in total. As a result of NTs, we isolated 83 PVs and 22 nonpolioviruses (Table 1).
The seasonal distribution of PVs and NPEVs during the study is presented in Fig. 1. After reisolation of a single PV serotype by an NT, we conducted sequence analysis of the PV capsid VP1 gene. The serotype frequencies of PV-positive isolates were 23 type 1 (PV1), 38 type 2 (PV2), and 22 type 3 (PV3) isolates (Table 1). Sequence analysis of the full-length VP1 region showed that all isolates were Sabin-like, with 1% nucleotide divergence from Sabin 1 (S1) and Sabin 3 (S3) and 0.6% divergence from Sabin 2 (S2) (30). As shown in Table 1, the frequency of PV-positive isolates obtained at plant T was >1.5 times higher than that obtained at plant Y. The distribution of each serotype showed a similar trend. These distributions mirrored the municipal vaccination (OPV) period in each area for ∼2 to 3 months (Fig. 1). The frequency of isolated PVs was highest in 2010 and tended to gradually decrease in the successive months of the sampling period. The last isolations of PV were in May 2012 at plant T and in November 2011 at plant Y. Since September 2012, routine IPV immunization has been conducted nationwide. Thus, before the immunization transition, PVs originating from OPV had disappeared from sewage water at both plants. On the other hand, NPEVs were isolated during the research period.
Frequency of poliovirus (PV) and non-polio enterovirus (NPEV) isolation at two independent sewage plants. The frequencies and distributions of NPEV and poliovirus type 1, type 2, and type 3 isolated from plants T and Y over the 34-month collection period were compared. The double-headed arrows above each graph indicate scheduled vaccination periods in each area. The dotted vertical lines indicate the month and year when IPV was introduced in Japan. The first y axis (left) indicates the frequency of PV isolation for each month, and the second y axis (right) indicates the frequency of NPEV isolation for each month. Sample collection was conducted from April 2010 to January 2013 (x axis).
Nucleotide substitutions in the VP1 region.The use of primers specific for the VP1 region of each PV serotype revealed mutations in the viral capsid sequence of the PV isolates relative to the Sabin vaccine strains. Of the 83 PV isolates obtained during this study, 69 (83%) showed at least one mutation; two PV3 variants had six nucleotide mutations in VP1 (Fig. 2).
Distribution of nucleotide substitutions in the VP1 regions of the isolated strains of poliovirus (PV). The nucleotide divergence of VP1 between isolates and reference vaccine strains (Sabin 1 [GenBank accession number AY082688], Sabin 2 [accession number AY082679], and Sabin 3 [accession number AY082683]) is shown on the x axis. The number of isolates with each mutation is shown on the y axis.
With regard to the mutations detected in the VP1 region, we closely examined the nonsynonymous mutations affecting the codons of amino acid residues. Many of these mutations are “attenuation markers” responsible for the attenuation phenotype of OPV strains (4). Several groups reported the following markers in the VP1 region of the Sabin vaccine strains: T106 and F134 in S1 (31), I143 in S2 (32), and T6 in S3 (33, 34). In the VP1 sequence of the 23 S1 isolates, we found 12 isolates with an amino acid substitution at T106 (52%) (Fig. 3). All these mutations exhibited nonsynonymous T→A/S mutations at residue 106 from the reference S1 strain. In contrast, no modified codon was found in the F134-coding region of the S1 isolates. Among the 38 S2 isolates, we found 20 isolates with a mutation at I143 (53%). These mutations also included three types of nonsynonymous substitutions (I143→T/V/N) affected by the first and second nucleotide modifications of this codon. On the other hand, the OPV Sabin 3 strain has been known to contain mutations at the second codon of the sixth amino acid position of VP1, even during manufacturing. In fact, all PV3 isolates had isoleucine at the sixth position, unlike the original Sabin 3 strain, which had threonine. In this study, we describe an amino acid change as a mutation. In summary, molecular analysis of the isolated PV serotypes revealed that several isolates contained mutations affecting amino acid residues of known attenuation markers of VP1.
Nucleotide substitutions in the VP1 region. The column above each nucleotide indicates the percentage of all PVs of an isolated serotype during the collection period that featured a mutation at this position. The circle below each codon shows the amino acid substitution resulting from the mutation. Capital letters below the nucleotides and in the circles represent conventional abbreviations for amino acids. The asterisk indicates that the OPV Sabin 3 strain originally contained a mutation at the second codon of the sixth amino acid of VP1, known as the “mixed-base position.”
DISCUSSION
Environmental surveillance of PVs is critical for maintaining polio-free areas and for working toward the global eradication of polio. We present the results of a 34-month environmental virus surveillance study conducted at two independent sewage disposal plants located in urban and rural areas of the northern Kyushu district, Japan. In total, 83 Sabin-like PVs were isolated from wastewater samples collected monthly from April 2010 to January 2013. The surrounding areas of each plant had unique features, including differences in sewage line coverage and infant age distribution, despite the similar population sizes (Table 1). We sought to determine the frequency of PV isolation from wastewater around the OPV-to-IPV transition period in those communities.
In Japan, the rate of usage of disposable diapers is quite high (∼80% to 90%). If disposable diapers containing stool from OPV-immunized children are properly treated and discarded, the frequency of PV detection in sewage water would presumably be quite low. However, PV was detected in wastewater sampled from two areas (disposal plants T and Y) with scheduled OPV immunizations and different infant age distributions. This finding suggests that either stool from OPV-immunized children flowed into the sewage line and contaminated the water or intrafamilial OPV transmission led to PV excretion into the sewage water. In fact, several reports have indicated intrafamilial and interfamilial spread of PV, including the vaccine strain and NPEVs (35–37).
Because of the high rate of usage of disposable diapers, PVs from OPV-immunized children may be transmitted by familial contacts, such as siblings and parents, or contacts at school facilities, such as kindergarteners (38). However, we also anticipate that NPEVs are widely transmitted among family members. In Toyama, Japan, a correlation between the high frequency of echovirus type 13 (E13) isolation from environmental water in the summer and seroconversion against E13 between preoutbreak and postoutbreak periods, regardless of age, was established (39). These studies suggest that we should expect differences in the frequencies of NPEV between the two plants. On the other hand, we showed that the frequency of PV was clearly related to the OPV vaccination period.
Previous studies have reported a rapid decline in PV isolation around the OPV-to-IPV transition period, with a disappearance of PV vaccine strains from wastewater within 2 to 3 months after the cessation of OPV administration (40–43). In contrast, our study showed the disappearance of Sabin-like PVs from the environment before OPV immunization had ceased (Fig. 1). PVs have not been detected in wastewater from plants T and Y since June 2012 and December 2011, respectively. After the announcement that would replace OPV for routine immunization in Japan (May 2011), nationwide coverage by routine OPV immunization declined until IPV was introduced in September 2012, mostly because of public concerns about VAPP (11, 12). The rate of VAPP is assumed to 1 per 2 million to 5 million inoculations in Japan. However, this rate might not be low for parents who have children who should have received the polio vaccine in this period. From now on, it will be very important to monitor the decline and disappearance of VAPP cases after OPV cessation, as demonstrated in the United States (6).
The analysis of the viral capsid VP1 sequence showed that all PV isolates were Sabin-like vaccine strains. Many isolates contained mutations in the VP1 sequence. However, major deviations from the attenuated Sabin strains were relatively rare, and there was no significant difference in mutations between the virus serotypes. Importantly, these data show that no VDPV has emerged and circulated in the sampling sites over the course of this study.
The PV1 isolates also presented several mutations affecting codons and amino acid residues, known as attenuation markers, in the VP1 region of all PV serotypes. In the Sabin 1-like isolates, a T→A amino acid mutation at position 106 was the most frequent mutation. This mutation introduces a residue found in the Mahoney strain, a neurovirulent parent of the Sabin 1 strain. With regard to the Sabin 2-like isolates, I143 was mutated (I→T/V/N) in 53% of the isolates. All Sabin 3-like strains isolated in this study had isoleucine at the sixth amino acid position of VP1, unlike the original Sabin 3 strain, which had threonine at this position. It is well known that the Sabin 3 OPV originally had mixed nucleotides at the second codon of the relevant position, even during manufacturing (44), but it is difficult for us to know at what stage the mutations occurred. Also, our study intended to show not the virulence and replication ability of the isolates but rather only the numbers of VP1 mutations for refuting the emergence of VDPV in the research field. This is one of the limitation points of this study. Besides this, the strains were isolated from sewage water and not from patients and other healthy persons. This is why we can only speculate on when the mutations occurred and the effect of VP1 mutations on the virulence of the isolates. More information may be obtained when we sequence the genomes of the isolates and compare them with known markers of other regions. However, our results indicate that continuous surveillance of nucleotide substitutions is necessary to monitor the emergence of VDPVs and other mutated strains.
In conclusion, we offer an environmental surveillance strategy that can detect the excretion of PVs from OPV-immunized populations with high sensitivity. Needless to say, the transition from OPV to IPV is an essential step toward the global eradication of PV. However, the transition from OPV to IPV should be carefully orchestrated before OPV cessation to maintain immunity coverage by OPV. Even if the IPV transition is successful, the risk of PV infection for a susceptible population must be closely monitored with a high-quality surveillance system, because PV can be silently transmitted and its viral genome can mutate during replication in the human gut, partly because of insufficient mucosal immunity (45, 46). Under these circumstances, environmental surveillance of PV plays a key role not only in monitoring the importation of wild-type PV from areas where the disease is endemic but also in preparation against emerging VDPVs before the global cessation of OPV at the final stage of polio eradication.
ACKNOWLEDGMENTS
This study was supported by a grant for Research on Emerging and Re-Emerging Infectious Diseases from the Ministry of Health, Labor and Welfare of Japan.
We appreciate the technical suggestions of Takenori Takizawa and Masae Iwai-Itamochi from the Toyama Institute of Health and the critical comments of Hiroyuki Shimizu and Chikako Kataoka from the National Institute of Infectious Diseases, Japan.
FOOTNOTES
- Received 28 October 2014.
- Accepted 20 December 2014.
- Accepted manuscript posted online 2 January 2015.
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
The authors have paid a fee to allow immediate free access to this article.