- Enteroviruses comprise polioviruses, coxsackieviruses, and echoviruses
- cause an estimated 10-15 million of more illnesses every year alone in the U.S.
- omnipresent in water including tap water
- higher prevalence in late summer and early fall
- transmitted via the fecal-oral route
- retain their infectivity longer when associated with solids
Enteroviruses are a large group within the enteric viruses. More specifically, they belong to the family of picornaviridae
. They are small-sized (20-30 nm, the term ‘pico’ historically indicated a ‘small’ size), non-enveloped, single-stranded RNA viruses with an icosahedral symmetry that enter the body through the gastrointestinal tract. The name “enteroviruses” attributes to the fact that this group of viruses thrives in the gastroenteric tract of humans and animals. The group comprises poliovirus (three serotypes), coxsackie (23 serotypes A and 6 serotypes B) and echovirus (31 serotypes) as well as a few ungrouped enteroviruses ( Health Canada 2004
). The group contains both human and animal pathogens. They are the second most common viral infectious agent in the world (Sawyer M 2002), and cause 30-50 million infections per year in the U.S. ( Gregory et al. 2006
). Whereas poliovirus is well known for causing poliomyelitis that damages the nerve cells that control muscle movement and can cause paralysis, coxsackie viruses are often causative agents of respiratory system infections, gastroenteritis, heart and other diseases. Echovirus infections are normally less dangerous and are associated with common cold symptoms and respiratory diseases as well as a variety of non-specific viral disease symptoms. Whereas 70% of coxsackie and echoviruses are associated with human disease, 30% are linked with animal infections ( Fong and Lipp 2005
). Interestingly, chronic enterovirus infections of the stomach have been associated with possible chronic fatigue syndrome ( Chia and Chia 2008
) and insulin-dependent diabetes mellitus ( Haverkos et al. 2003
; Kawashima et al. 2004
). Enteroviruses are transmitted via the fecal-oral route. According to a factsheet about enteroviruses by CDC (Center for Disease Control), they cause an estimated 10-15 million of more illnesses every year alone in the United States (http://www.cdc.gov/ncidod/dpd/healthywater/factsheets/enterovirus.htm
). CDC recommends removing enteroviruses by boiling the water to be consumed for one min, at high altitudes for 3 min to inactivate the viruses. Point of use filters do not remove enteroviruses due their small size. The efficacy of disinfection depends on the resistance of the particular virus and serotype, relatively high resistance has been reported.
Enteroviruses are omnipresent in water including tap water ( Lee and Kim 2002
; Lee et al. 2005
), river water ( Lee et al. 2005
; Skraber et al. 2004
), groundwater ( Abbaszadegan et al. 1999
; Fout et al. 2003
) and well water ( Jean et al. 2006
), recreational waters ( Rose et al. 1987
), estuarine and sea water ( Finance et al. 1982
). Infections are most likely to occur during the summer and early fall coinciding with peak recreational activities ( Fong and Lipp 2005
). Viruses originate from contamination with feces from infected individuals that enter the water for example by sewage overflows or broken sewage systems or water runoff. A study in Taiwan correlated a higher probability of enterovirus infections caused by well water contamination with heavy rainfall ( Jean et al. 2006
). The rain was suspected to flush enteroviruses into aquifers. Especially in raw sewage very high concentrations can be reached with concentrations varying between from a few hundred t o more than 100,000 per liter in some regions of the world ( AWWA 2006
). The average concentration in raw sewage in the U.S. is estimated to be 7,000 per liter with higher prevalence in late summer and fall ( AWWA 2006
). The enteroviruses isolated most often from water are coxsackiviruses. Polioviruses are often isolated at the same time when oral vaccinations are administered ( Iwai et al. 2007
). Many isolates are identical with vaccine strains. Examples of studies examining enterovirus occurrence and abundance are given in the following:
- A nine-year study of different viruses in influent and effluent water of a WWTP in Milwaukee revealed maximum enterovirus titers of up to 3,300 enterovirus MPN per liter for WWTP influent. Testing was performed using a modified U.S. EPA organic flocculation cell culture procedure. A strong seasonality for WWTP influent samples was observed: Most isolates were found during the months of July through October coinciding with the frequency of enterovirus-caused clinical cases. The same serotypes causing clinical problems were also found in WWTP influent every year. Viral prevalence and titers were dramatically lower in the WWTP effluent: Whereas 98% (n=107) of influent samples tested positive for culturable viruses, only 30% of effluent samples (32 samples out of 107 total samples) tested positive for culturable viruses with much lower titers with a strong prevalence of reoviruses and adenoviruses. Only 4 of the 32 positive effluent samples contained enteroviruses ( Sedmak et al. 2005).
- Gantzer et al. tested three types of treated wastewater in France for the presence of infectious enteroviruses, enterovirus genome, somatic coliphages, and Bacteroides fragilis phages. Quantification was achieved by cell culturing for enteroviruses and plaque formation (on E. coli or B. fragilis host cells) for the phages. Enterovirus genomes were detected by RT-PCR. Wastewater was subjected to secondary treatment of 8 to 12 hours (A) or for 30 treatment (B1), or to both secondary and tertiary treatment (B2). The results from 16 samples for each treatment level are summarized in the following table:
As the level of treatment increased, the concentrations of all analyzed biological agents decreased. The percentage of samples testing positive for enterovirus genomes was always higher than the number of samples testing positive for infectious enteroviruses. It was concluded that B. fragilis phages might be good indicators of enterovirus contamination as the fluctuations in their concentrations correlated well ( Gantzer et al. 1998).
- A study in The Netherlands examined concentrations for enteric viruses in different water samples. The concentrations in raw sewage ranged from 99 to 1,141 pfu per liter and in treated sewage from 3 to 424 pfu per liter in winter months. The average calculated removal rate by sewage treatment was 1.4 (0.7 to 1.8) log10 units. In two rivers under the influence of (un)treated sewage, the Maas and Waal river, the estimated enterovirus concentrations were in the range between 0.1 and 3 pfu per liter ( Lodder and Husman 2005).
- A German study examined viral loads of the waters located downstream of a WWTP over one year. In total, 9 samples were taken from the WWTP influent and 123 samples from the waters downstream (river, pipeline, and lake). Among enteroviruses, astroviruses, noroviruses, rotaviruses, adenoviruses, and hepatitis A, enteroviruses were the most common viral group detected (using nested RT-PCR). Detection rates decreased as the distance from the WWTP influent increased: Whereas 89% of samples tested positive at the wastewater influent, 76% were positive at the river directly downstream, 33% positive after transfer through a pipeline system, and 29% positive at the shore of a receiving lake. The decrease was explained by dilution effects. A total of 18 PCR-positive samples were examined for the presence of infectious viruses. Three samples from the site after the pipeline passage tested positive by cell culture, all other downstream samples tested negative. Despite the high detection rate by nested PCR, qPCR suggested relatively low enterovirus genome loads in the samples. Analysis of five selected samples from the river suggested an average of 6.4 x 102 enterovius genomes per liter ( Pusch et al. 2005).
- The presence of enterovirus genomes is more common than the presence of infectious enterovirus particles. A study examining four French rivers monthly or semimonthly over one year, enterovirus genomes were detected in 88% (n=68) samples (20 liters each) using RT-PCR, whereas infectious viruses were only isolated in only 3% of these samples using cell culture ( Hot et al. 2003).
- A Japanese study by Tani et al. monitored monthly levels of different viruses during a 63-month period in an urban river. Reoviruses were isolated most frequently throughout the entire time with levels peaking in colder winter months. For enteroviruses a seasonal distribution was observed with peak levels in summer months. Monthly levels ranged from 0-190 pfu per liter with an average of 40.6 pfu per liter. A changing pattern of serotype prevalence was observed for coxsackie B viruses and echoviruses, whereas this was not the case for polioviruses which remained nearly unchanged during the 5 year period. Polioviruses were isolated during the administration of vaccine ( Tani et al. 1995).
- Applying an optimized filtration-RT-qPCR method for small volumes of natural waters (one liter), enteroviruses were found in two freshwater creeks that drain into Santa Monica Bay, California. One out of 10 samples from the relatively pristine Topanga mountain creek tested positive with one genome per ml, and 2 out of 7 samples from the concrete-lined urban storm drain (Ballona Creek) tested positive with 10 to 25 genomes per ml ( Fuhrman et al. 2005).
- Vivier et al. studied the presence of infectious enteroviruses in two South African drinking water supplies over a period of one year ( Vivier et al. 2004). Drinking water samples originated from surface water sources with acceptable quality after treatment and disinfection according to international standards. Detection was performed using an integrated cell culture/nested RT-PCR approach, followed by restriction analysis for molecular typing. Enteroviruses were detected in 11% and 16% of the samples from two drinking water plants. With 88% the coxsackie B viruses represented the majority of the detected viral particles, the remaining 12% were polioviruses. Serotypes B3 and B5 were the most abundant of the coxsackie B viruses. A follow up study with an extended range of water sources found enteroviruses in 42% (n=100) of sewage, 19% (n=850) of drinking water, 28% (n=60) of river water, 27% (n=197) of dam/spring water, and 25% (n=108) of borehole samples. Restriction typing again suggested a high prevalence of coxsackie B viruses (71% of typed PCR products), followed by coxsackie A viruses (28%) and echoviruses (6%) ( Ehlers et al. 2005).
- A Korean study detected enterovirus in tap water at levels sufficient to cause disease using ICC—multiplex-nested PCR (for simultaneous detection of enteroviruses and adenoviruses). Of 50 tap water samples taken at the distal point of the distribution system in Seoul metropolis, 30% tested positive for enteroviruses by ICC-PCR or cell culture (and 58% for enterovirus and adenovirus combined). The tap water volume processed was between 1,000 and 3,000 liters. Concentrations of both viruses combined were in the range from 1.2 to 28.9 MPNIU per 1,000 l tap water (determined by ICC-PCR). When analyzing surface water from the Han river at three intake sites for tap water for the Seoul metropolitan area (70 – 300 liters), around 46% of 69 samples were positive for enterovirus (by ICC-PCR or cell culture). ICC-PCR suggested viral particle concentrations (for both viruses) in the range from 0.6 to 84.0 MPNIU per 100 l surface water ( Lee et al. 2005).
- When testing groundwater from different geographical locations in the U.S. for enteric viruses, 30.1% of samples (40 out of 133 samples) tested positive for enterovirus RNA by RT-PCR. Out of the 40 PCR-positive samples, six also showed cytopathogenic effects in cell culture ( Abbaszadegan et al. 1999).
- Enteroviruses were found in well water from a confined deep sandstone aquifer serving as a public water supply. Seven of 30 samples (collected over a period of 15 months) tested positive for human enteroviruses by RT-PCR. Cell culture revealed one of these samples being positive for infectious echovirus 18 ( Borchardt et al. 2007).
- In a study looking at enteroviral presence in the estuary of Galveston Bay, 27% (n=103) of samples tested positive for enteroviruses. Enteroviruses were found to be mostly associated with solids: 72% of samples were positive for suspended solids (n=18), 47% for fluffy sediments (n=15), 6% for sediments (n=35). Only 14% of water samples (n=35) seemed to contain enteroviruses. Poliovirus 2 was recovered most often (39% of isolates), followed by poliovirus 1 (18%), coxsackievirus B-4 (14%), coxsackievirus B3 (10%), echovirus 7 (7%), echovirus 29 (7%), and poliovirus 3 (3%). Virus concentrations were lowest for water (in the range ranged from 3 to 12 PFU per 100 gallons (379 liters) of water) and highest for fluffy sediments (in the range from 39 to 398 PFU per 1,000 g) ( Rao et al. 1984). The finding were discussed in the context that viruses remained adsorbed to marine sediments over a wide range of temperature, pH, and salinity conditions (LaBelle and Gerba 1979). Viruses accumulated by shellfish might be mostly particle-associated.
- Gersberg et al. studied enterovirus levels in water from recreational beaches near San Diego, California, which are under the influence of untreated sewage from the Tijuana river. The river receives sewage from the Tijuana region, Mexico. A total of 20 samples, taken over a 2-year period, were analyzed using RT-qPCR. Enteroviral genomes were found in 93% of wet-weather samples with concentrations ranging from 7 to 4,417 PDUs per liter. In dry-weather samples, the levels were below the detection threshold. All samples were also analyzed for hepatitis A virus whose levels showed the same tendency (see corresponding section). The much higher prevalence during the wet-weather season with increased rain fall was attributed to the inadequate sewage treatment infrastructure in the Tijuana region resulting in system overload ( Gersberg et al. 2006).
- Using RT-qPCR, Rose et al. found enterovirus genomes in all eight samples studied of Lido beach in Venice in concentrations between 2 and 71 copies per liter (average app. 16 copies per liter). In the Grand Canal, concentrations in 7 samples ranged between non-detectable to 1614 genomes per liter (average 270 copies per liter) and in two interior canal samples 164 and 51 genomes per liter were detected. The authors point out that these results are not surprising considering the high degree of sewage contamination of these waters ( Rose et al. 2006).
The degree of resistance towards disinfection varies greatly among enteric viruses. Research from Payment et al. has shown that some enteroviruses can survive of several hours in chlorine-treated water ( Payment et al. 1985
). Infectious viruses were still detected after exposure to residual chlorine concentrations of 0.1 mg per liter for 16 hours. This explains the fact that enteroviruses can still be found in finished tap water, although in small numbers. The residual infectious fraction after 16 hours of exposure is believed to be less than 0.001 % for most viruses ( Payment et al. 1985
). In general, coxsackieviruses tend to be most resistant to disinfection and polioviruses least. However, a low degree of genetic resistance of purified viral particles to a disinfectant does not necessarily indicate a high degree of inactivation. Among the factors influencing survival of viruses, aggregation and the presence of particulate or organic matter can offer significant protection ( Payment et al. 1985
). Enterovirus adsorption to natural organic and inorganic solids in various types of natural waters was reported significant over a wide range of pH and various concentrations and species of metal cations ( Schaub and Sagik 1975
). Cations like Ca2+
in general favor viral adsorption to solids. Clay-adsorbed viruses retained infectivity. In a study trying to determine the degree of adsorption of enteric viruses (poliovirus 1, coxsacievirus B3, echovirus 7, and rotavirus SA-11) to estuarine sediment, more than 99% of viruses were found to be adsorbed ( LaBelle and Gerba 1979
). It is essential to consider that in natural water viruses can be efficiently protected from disinfectants and heat in the presence of organic particles and sediments ( Fong and Lipp 2005
; Liew and Gerba 1980
). Adsorption of viruses to particulates alters survival patterns and can result in increased persistence of infectivity under various conditions compared to data obtained from pure suspensions in a laboratory setting. Also association to cells can provide increased resistance. Whereas a suspension of free poliovirus 1 and coxsackievirus A9 were completely inactivated by 0.081 mg of ozone per liter within 10 sec in a continuous-flow ozonation system, ozone dosages of 4.06 and 4.68 mg per liter for 30 sec resulted in only a 3-4 log10 unit reduction of the viruses attached to HEp-2 cells under the same conditions ( Emerson et al. 1982
). Selected publications addressing susceptibility to disinfection are outlined in the following:
- Payment et al. studied the survival of different coxsackie- and polioviruses when exposed to an initial concentration of app. 0.4 mg of free chlorine per liter for up to 1,000 min to simulate water treatment conditions in water reservoirs or distribution systems ( Payment et al. 1985). Infectious viral particles were assayed in a plaque test using BGM cells. Environmental isolates originating from water disinfected with chlorine tended to be more resistant than lab strains although the differences were small. One coxsackievirus B5 strain isolated from raw sewage was found to be most resistant with 83%, 70%, and 22% of the initial population surviving after 1, 10, and 100 min of contact time, respectively. Even after 1,000 min of exposure 0.079 % of the viruses were still infectious. Coxsackievirus B5 was also reported earlier to have a high resistance to chlorine ( Engelbrecht et al. 1980). CoxsacievirusB5 was followed in regard to chlorine resistance by coxsackievirus B4, and poliovirus 1, 2, and 3. Among the viruses tested, the least resistant poliovirus 3 strain was reduced to 0.03% of the original number after only 1 min of exposure. Longer exposure times resulted in reduction to <0.003% ( Payment et al. 1985).
- Complete loss of poliovirus 1 infectivity was achieved by exposure to 5 mg L-1 of chlorine dioxide for 3 min. Detection of viral genomes measured by RT-qPCR overestimated the infectious risk and was in disagreement with culture results. However, genome degradation and loss of infectivity were dose-dependent ( Simonet and Gantzer 2006).
- Gantzer et al. studied the survival of infectious coxsackievirus B3 at different temperatures compared to the intactness of the viral genome. Infectious particles were undetectable after incubation at 55°C for 15 min, while the viral genomes were still detected by RT semi-nested PCR after exposure to 95°C for 15 min. The faster loss of infectivity was assumed to be due to cleavage or conformational changes in the viral capsid while the genome was still intact ( Gantzer et al. 1996). In natural water, significant protection of viruses from heat inactivation by organic particles has to be considered. Liew and Gerba reported prolonged survival of poliovirus 1 and echovirus 1 in the presence of sediment ( Liew and Gerba 1980). The infectivity in artificial seawater containing sediment was at least 1 log10 unit higher at 24°C and 37°C at most time points compared to the infectivity of viruses in seawater alone.
- Lamont et al. studied inactivation of poliovirus 1a by pulsed UV-light. Viral particles suspended in 2 ml of PBS (concentration of app. 106 tissue culture infectious units per ml) were exposed to varying numbers of pulses emitted from a xenon flashlamp. Infectious particles were quantified using cell culture (MA104 cells). The authors reported a virus titre reduction of app. 4 log10 units after 10 pulses. No infectious polioviruses were detected after 25 pulses. In comparison, adenoviruses were more resistant needing 200 pulses to achieve a greater than 4 log10 unit reduction. Pulsed UV-light can be seen as an alternative to continuous-wave UV treatment ( Lamont et al. 2007).
- Solar disinfection: Haeselgrave et al. observed a total inactivation of infectious poliovirus type 2 at 40°C when subjected to simulated global solar irradiation (SODIS) of 850 Wm-2 with a steady reduction within the exposure time of 6 hours. Survival was measured by cell culture. At 25°C only a small decrease was observed for the first 2 hours, followed by a dramatic loss of viability by 4 hours. The data suggested that SODIS disinfection should achieve complete inactivation of the virus ( Heaselgrave et al. 2006).
- A dose-response relationship was demonstrated for ozone inactivation of poliovirus 1. A virus suspension was mixed with ozonated water in a fast-flow mixer. The mix was subsequently passed through a neutralizing solution. The ozone concentration ranged between 0.1 to 2 mg per liter. Inactivation of viruses occurred in two steps: In a first step lasting for 0.2 to 1.0 sec, 95 to 99% of the viruses were inactivated depending on the ozone concentration. The remaining infectious particles were inactivated in a second step which lasted for several minutes. A first order kinetic reaction was observed for viral inactivation ( Katzenelson et al. 1979).
- Poliovirus 1 infectivity was reduced by ozone by >4.7 log10 units within a contact time of 5 min. The initial ozone dose was 0.37 mg per liter (pH 7, 5°C) ( Shin and Sobsey 2003).
- A synergistic or additive effect of copper or silver ions in combination with chlorine was found. Electrolytically generated copper and silver ions (400 and 40 μg per L) led to an inactivation of poliovirus type 1 at a rate of 0.0006 log10 units per min, compared to 0.036 log10 units per min by exposure to 0.3 mg/L free chlorine. The inactivation rate in the presence of copper or silver and chlorine in combination resulted in significantly faster inactivation than exposure either to metals or chlorine alone. Poliovirus was app. 10 times more resistant to the disinfectants than coliphages MS-2 ( Yahya et al. 1992).
Enteroviruses are quite resistant to environmental stress and are stable under a wide range of pH conditions. Stability for 1 to 3 hours at pH 3 to 5 and for minutes at pH 10 to 11 has been reported ( AWWA 2006
). They can survive drinking water treatment and have been detected in treated drinking water apparently free of coliform bacteria ( Health Canada 2004
). As an example, viable enteroviruses (predominantly coxsackie B viruses) were detected in 11% and 16% of chlorinated drinking water samples from two drinking water plants in South Africa ( Vivier et al. 2004
). Temperature is seen as one of the principal factors affecting survival in the environment. The higher the temperature, the faster the inactivation of the virus (O’Brien and Newman 1977). Hurst et al. studied the long-term survival of coxsackievirus B3, echovirus 7, and poliovirus 1 in seeded surface water samples collected from five sites of physically different character (artificial lake, groundwater outlet pond, a large and a medium-sized river, and a suburban creek). The average rate of viral inactivation was reported to be 6.5-7.0 log10
units over 8 weeks at 22°C, 4-5 log10
units over 12 weeks at 1°C, and 0.4-0.8 log10
units over 12 weeks at -20°C. Hardness and conductivity, turbidity and suspended solids, and the number of generations of bacterial growth that a sample could support were determined as the apparent parameters influencing viral persistence ( Hurst et al. 1989
Viruses retain their infectivity longer when associated with solids. Adsorption to solids can offer protection from enzymes, UV degradation, and other degrading factors ( Fong and Lipp 2005). Rao et al. showed for an estuarine environment that polioviruses remained infectious for up to 19 days when bound to suspended estuarine sedimentary material (suspended solids and fluffy sediment), compared to only 6 days when freely suspended in seawater ( Rao et al. 1984). Survival was measured by cell culture. Another important factor affecting viral persistence in water is the presence of indigenous microorganisms whose presence normally leads to a faster degradation of enteroviruses (examples are presented below). On the other hand, indigenous organisms in the form of a biofilms might provide a sheltered environment and the biofilms might act as an attractive surface. Skraber et al. describes in a review an experiment performed by Hock and Botzenhart in which biofilms were spiked with 1.3 x 104 pfu/ml of poliovirus-1. Cell culture tests indicated persistence of infectious viral particles for 7 days with 2.8 pfu/cm2 recovered. RT-PCR allowed detection of viral RNA after 14 days, but not by 21 or 28 days ( Skraber et al. 2005). The review article points out that there is a chance that viruses might accumulate in biofilms from bypassing water over time. Biofilms might also protect viruses from disinfection.
Examples of selected studies evaluating enterovirus survival under different conditions are given in the following:
- Payment et al. studied the elimination of viruses and indicator bacteria at each step of treatment during the preparation of drinking water by cell culture methods. Seven utilities were sampled, treatment included chlorination, sedimentation, filtration, and ozonation. Enteroviruses were found to persist most during treatment as all of the viruses isolated from treated waters were enteroviruses (poliovirus types 1, 2, and 3; coxsackievirus types B3, B4, and B5; echovirus type 7; and untyped picornaviruses). Interestingly, no reoviruses could be detected in the finished water, although they accounted for high numbers in the raw water. The presence of enteroviruses in finished water cannot be explained solely by their resistance to disinfection. Only isolates of coxsackieviruses showed adequate resistance against ambient chlorine concentrations (0.5 mg free chlorine per liter) to explain the survival, whereas poliovirus numbers were reduced more efficiently at the same concentration. The latter might have been protected from disinfection by some other mechanism ( Payment et al. 1985).
- Skraber et al. compared survival rates of Poliovirus-1 particles, thermotolerant coliforms and somatic coliphages in artificially contaminated river water. Survival was monitored over two weeks. River water samples were taken in summer and winter before spiking and incubation at different temperatures (4°C, 18°C, and 25°C). Poliovirus-1 particles (measured by cell culture) survived 1.5-fold longer than culturable thermotolerant coliforms, but 2-6-fold shorter than somatic coliphages. The Poliovirus-1 genome (measured by RT-qPCR) persisted about 2-fold longer than the infectious Poliovirus-1 particles. It was concluded that thermotolerant coliform counts might underestimate enteroviral numbers in river water. Somatic coliphages and viral genomes would be more suitable as indicators of enterovirus survival ( Skraber et al. 2004).
- O’Brien studied inactivation rates of enteroviruses (poliovirus 1 and 3; coxsackieviruses A-13 and B1) in situ in the Rio Grande river in New Mexico using membrane dialysis chambers. Infectious particles were quantified by infecting HeLa cell monolayers. Coxsackievirus B-1 and poliovirus 1 were the most stable serotypes. Temperature was concluded to be a decisive factor for loss of infectivity. At water temperatures between 23 and 27°C typical 1-log reductions of infectivity required 25 h for poliovirus 1, 19 hours for poliovirus 3, and 7 hours for coxsackievirus A-13. Lowering the temperature led to a gradual slowdown of virus inactivation. At river temperatures between 4 and 8°C mean 1-log reduction times increased to 46 hours for poliovirus 1 and 58 hours for coxsackievirus B-1. Inactivation was presumably due to damage of the viral RNA genome as inactivation did not lead to loss of capsid proteins and did not affect the ability of viruses to adsorb to host cells. Inactivation in autoclaved river water was substantially slower in autoclaved river water compared to raw water or filter-sterilized (0.22 μm) water (O’Brien and Newman 1977).
- Temperature was also found to be the critical factor for enterovirus survival in estuarine and marine waters rather than salinity. Lo et al. studied the survival of poliovirus 1, echovirus 6, and coxsackievirus B-5 under controlled laboratory conditions and in situ using cell culture. At all conditions tested, coxsackievirus B5 was most stable. Infectivity dropped faster under lab conditions (with artificially prepared estuarine and marine waters) compared to in situ experiments. Under laboratory conditions, infectious coxsackievirus particles could be detected up to 8-10 weeks at 25°C, 40-53 weeks at 15°C, and at least 53 weeks at 4°C at salinity values between 10 and 34 parts per thousand. Echovirus 6 had an intermediate stability, poliovirus 1 was least stable with infectious particles being detectable up to 46 weeks at 4?C each. Under in situ conditions, survival was greatly enhanced in winter months, compared to summer months. For Poliovirus 1 plaque forming units were counted for up to 27 and 65 days for summer and winter months in ocean water, respectively, compared to 31 and 70 days for echovirus 6. Infectious coxsackievirus B-5 particles could be detected for up to 7 weeks in summer and several months in winter. In the winter infectivity dropped only by 2 log10 units over and 80-day period. These findings have consequences for sewage discharge into marine environments ( Lo et al. 1976).
- Poliovirus type 1 and coxsackievirus A9 were found to be more rapidly inactivated in a lake than in sterile lake water. Submersion of virus suspensions (initial titer of >107 PFU, suspension filled in a dialysis bag) in a lake at a depth of 1 m resulted in a 5 log10 reduction of infectivity in 9 days and 21 days for coxsackievirus and poliovirus, respectively. Infectivity was measured by a plaque assay. In sterile lake water, infectivity of the two viruses only decreased by app. 1.3 and 2.4 log units in the same time period. Loss of infectivity occurred significantly faster than loss of a 14C label (viruses were labeled with 14C-leucine after propagation) suggesting that biodegradation of coat proteins is not the primary cause of loss of infectivity ( Herrmann et al. 1974).
- Yates et al. studied the survival of poliovirus 1 and echovirus1 in groundwater collected from 11 sites throughout the United States. Aliquots of 50 ml groundwater were seeded with viruses to a final concentration between 104 to 106 PFU per ml and incubated at the in situ groundwater temperature of the collection site. Decay rates for the two enteroviruses were similar and were in the range between 0.035 to 0.676 log10 PFU per day. The mean decay rate for poliovirus 1 was found to be 0.1615 log10 units per day. This value was correlated with poliovirus 1 decay rates from previous studies, which were 0.21 (Keswick et al. 1980) and 0.0456 ( Bitton et al. 1983) log10 units per day. Decay rates for MS-2 coliphage were not found significantly different from the ones for enteroviruses ( Yates et al. 1985).
- Biziagos reported only minor loss of poliovirus 1 infectivity in mineral water stored at 4°C for one year. At room temperature, infectious particles were not detected any more after 300 days (HAV at the same time under the same conditions was still infectious after 300 days). Infectivity decreased 1.55 log10 units in 84 days at 23±3 °C. A decrease of 0.0193 log per day in this study was compared to decay rates between 0.0456 to 0.21 log per day in groundwater and 0.087 log per day in laker water. Times required to obtain a 3 and 5 log10 decrease of infectivity were 211 and 261 days, respectively ( Biziagos et al. 1988). A previous study reported a 5 log decrease of infectivity after 84 days at 25°C ( Sobsey et al. 1988).
- A thorough study on the influence of groundwater characteristics on the survivial of seeded poliovirus 1, coxsackievirus B1, E. coli, and bacteriophage MS2 was performed by Gordon and Toze. Changes in MS2 and E. coli numbers were determined by cultivation, while the enterovirus numbers were determined by RT-qPCR. The most influential factor affecting viral persistence was found to be the presence of indigenous microorganisms. Abiotic factors like temperature, dissolved oxygen, and nutrient levels were believed to indirectly affect viral survival by influencing the activity of the groundwater flora. When testing 15 and 28°C, neither temperature had much influence on the survival of the viruses or E. coli in the absence of microorganisms, the decay was reported to be very slow. Only for E. coli, decay increased after day 21 at both temperatures. For the viruses the decay was continuous and was a little slower at 15°C. When the indigenous groundwater flora was present, the decay rates were much faster for all four microorganisms. For the two enteroviruses the decay was staged with a rapid loss of the first short stage and a much slower loss after day 21. The mean removal times (times required for a one log reduction) for the two enteroviruses are shown in the following table (the table represents a shorter version of the table in the original publication):
In accordance with the temperature data, dissolved oxygen (aerobic or anoxic) and nutrient levels tended to have a significant effect on decay rates of the four microorganisms studies only when the indigenous groundwater flora was present. The study all in all suggests a significant influence of the indigenous microbial flora on the decay rates of the microorganisms studied ( Gordon and Toze 2003).
Although enteroviruses are a significant public health concern in developing countries, little is known about their transmission by water ( Health Canada 2004
). The large number of serotypes and the unknown composition of serotypes in a water sample do not allow to establish dose-response relationships for enteroviruses as a group. For coxsackievirus A21 for example an infectious dose of less than 18 infectious units was reported by inhalation (http://www.phac-aspc.gc.ca/msds-ftss/msds44e.html
). For poliovirus 1, the orally administered dose to infect 50, 10, and 1% of 32 infants were estimated to be 72, 39, and 20 tissue culture infective doses, respectively ( Minor et al. 1981
). A study with 149 human volunteers concluded that the echovirus 12 dose required for infection of 50% of healthy adults with undetectable serum antibody was 919 PFU ( Schiff et al. 1984
). Statistical analysis predicted 17 PFU as a 1% infectious dose for humans. However, one plaque forming unit can comprise multiple infectious particles which adhere together. A particle to PFU ratio between 30 and 1000 was suggested for poliovirus and other enteroviruses ( Racaniello VR 2001
). In the echovirus 12 study, electron microscopic evaluation of the virus preparation suggested that 17 PFU would correspond to about 700 infectious particles. A previous infection with this virus did not provide lasting protection again reinfection. In general it is assumed for risk assessment for water contact that one infectious unit is capable of successful infection in a susceptible person ( Percival et al. 2004
The detection of enteroviruses in cell culture has recently experienced some great technological advances which are addressed briefly. Cultivation-based plaque methods have been ranked in terms of efficiency by Mocé-Llivina et al.: Double-layer plaque assays ≥ suspended-cell plaque assay > counting cytopathogenic virus adsorbed to cellulose nitrate membrane filters (VIRADEN) ≥ most probable number of cytopathogenic units > monolayer plaque assay ( Mocé-Llivina et al. 2004
). Two of those techniques are described in more detail in the following:
Although culture analysis is still the gold standard, the high time demand of 5-10 days does not allow preventive measures in case of positive results. The use of fast molecular assays would significantly shorten analysis time with the drawback that assays simply based on PCR amplification of nucleic acids cannot distinguish between infectious and non-infectious particles. When comparing cytopathogenic enterovirus numbers obtained by cell culture with the number of genomes determined by RT-qPCR in activated sludge, no correlation was found ( Pusch et al. 2005). Although this certainly depends on the sample and the time which passed from contamination with infectious particles to detection, PCR-based data have to be interpreted with a lot of care. RT-qPCR values tend to be considerably higher than the numbers of CPE. One way to minimize overestimation is to target highly susceptible regions of the viral genome. When studying the degradation of poliovirus 1 genome by chlorine dioxide by RT-PCR, Simonet and Gantzer found that the preferential sites of action of this disinfectant were the untranslated regions at the 5’- and 3’-end ( Simonet and Gantzer 2006). Also the length of the targeted sequence should not be too small: Whereas a targeted fragment of 76 bp persisted throughout disinfection, a app. 7,000 bp fragment disappeared rapidly. The authors point out that RT-PCR overestimates infectious risk when short fragments are detected. However, even when targeting a long fragment, disagreement between signals from RT-qPCR and cell culture could not be eliminated.
If susceptible cell lines exist, the combination of molecular detection with cell culture remedies the disadvantage of indifferent detection of infectious and non-infectious particles and is still significantly faster than pure cell culture detection although it does not produce actual virus counts. When comparing ICC-PCR with conventional cell culture methods for detecting enteroviruses and adenoviruses in environmental samples, Greening et al. concluded that ICC-PCR was the preferred detection method for samples with high viral concentrations ( Greening et al. 2002). It was less successful, however, for samples with low viral titers or in the presence of toxic materials where direct RT-PCR or the suspended cell plaque assay showed higher sensitivities.
An overview over selected molecular detection assays with examples is given in the following:
- VIRADEN (Virus adsorption enumeration): VIRADEN is a modified plaque assay method which allows enumeration directly from membrane filters. In a first step viral particles are adsorbed to cellulose nitrate membrane filters from aqueous samples. Subsequently, the filter is carefully placed upside down on top of a cell suspension (e.g. BGM cells) which was spread before on a cell monolayer. Cells cover the entire surface of the filter and expand into the pores and penetrate them where they reach the adsorbed viruses. Viral particles infect the cells, replicate, and form countable plaques ( Papageorgiou et al. 2000). VIRADEN allows the use of different cell lines in the two cell layers. It can be applied to virus enumeration in tap water ( Papageorgiou et al. 2000), sewage ( Papaventsis et al. 2005), seawater and bathing water (Mocé-Llivina et al. 2005). The technique is especially well suited for viral enumeration in samples with significant numbers of enteroviruses like raw sewage and secondary sewage effluents ( Papaventsis et al. 2005).
- Double-layer plaque assay: This enterovirus quantification method combines a traditional monolayer plaque assay with a suspended-cell plaque assay. The monolayer is at the bottom surface of the culture plate, which is covered by a thin film (1-2 mm) of suspended cells in a semisolid layer. Viral counts were reporter to be one order of magnitude greater than counts obtained from monolayer assays. The method allows the use of two different cell lines in the two layers facilitating the recovery of a greater number and diversity of enteroviruses ( Mocé-Llivina et al. 2004)
- Multiplex RT-PCR: A multiplex assay was developed for simultaneous detection of polioviruses, coxsackieviruses, echoviruses and hepatitis A virus in water samples. Agarose gel visualization showed specific PCR products. Detection sensitivity of the multiplex reaction was similar to that of the monoplex RT-PCR which was 34 pfu for poliovirus, 21 pfu for coxsacievirus, 60 pfu for echovirus, and 105 TCID(50) for hepatitis A virus ( Li et al. 2002). A RT-nested PCR assay for simultaneous detection of enteroviruses and adenoviruses was developed by Cho HB and al. (Cho HB 2000). The sensitivity of the multiplex assay was found similar to the one of the monoplex PCR when tested on environmental water samples.
- RT-qPCR: A quantitative RT-PCR TaqMan assay based on previously published primers ( Monpoeho et al. 2000) was tested by Corless et al. for detection of enteroviruses ( Corless et al. 2002). The system was equally sensitive for? duplex detection of enteroviruses and parechoviruses. A two-step reaction with separate RT and PCR reactions was more sensitive than a one-step reaction by about 3 orders of magnitude and 100 fold more sensitive than cell culture. The assay was successfully tested on 22 enterovirus serotypes including coxsackieviruses A and B, echoviruses, and polioviruses. The assay was been tested with clinical samples.
- RT-qPCR with IAC: A quantitative TaqMan RT-PCR assay was developed targeting the 5’ untranslated region of enteroviruses. A synthetic target which was used to spike samples and which was reverse transcribed and amplified with the same primers served as a competitive internal positive control. Its inclusion allows for correction of virus concentrations in case of PCR inhibition. Samples which exhibited significant inhibition were diluted resulting in relief of inhibition. The assay was reported to detect app. 25 enteroviral genomes with a dynamic range of >3 log units. The time demand is around 5 hours ( Gregory et al. 2006).
- ICC-RT-(nested) PCR: Concentrated water with potential viruses are added on a monolayer of cells (typically BGM or BGMK), infectious viruses are allowed to infect cells for several days. After elimination of non-infectious viral particles through washing steps genomic DNA is isolated from the cell culture and analyzed for the presence of viral genomes using RT-nested PCR. The assay targets the 5’ untranslated region of the enterovirus genome and could detect very few viral particles. It was successfully applied to detect enterovirus in surface water samples ( Chapron et al. 2000). The same primers (originally published by Puig et al. 1994) were used in an ICC-RT-PCR DNA enyzme immunoassay by Skraber et al. for enterovirus detection in river water ( Skraber et al. 2004). Amplified DNA was analyzed using the DiaSorin DEIA kit.
- ICC multiplex nested RT-PCR: Lee et al. developed an integrated cell culture-multiplex-nested PCR method for simultaneous detection of enteroviruses and adenoviruses in environmental waters. A single cell line, buffalo green monkey kidney (BGMK) cells, was used for the two viruses. The protocol was tested on 69 surface water and 50 tap water samples. Results were compared with the ones obtained by a cell culture approach. For enteroviruses, both methods agreed completely and suggested presence of infectious particles in 46% of surface water and 30% of tap water samples. Good agreement was also found for adenoviruses, however more positive results were obtained with the ICC-PCR method than with cell culture alone. It was suggested that relatively long incubation times (10 days and more) may be required to detect fastidious viruses (such as adenoviruses) at low concentrations ( Lee et al. 2005).
- Reporter cell line-FRET: An efficient enterovirus detection system was developed based on genetically modified BGMK cells that express a hybrid fluorescent indicator. The indicator is composed of a cyan fluorescent protein (CFP) and a yellow fluorescent protein connected through a linker peptide. The two fluorophors act as a donor and acceptor undergoing FRET. The energy transfer neutralizes the CFP signal. Upon viral infection of the host cells, the linker peptide is cleaved by the 2A protease, encoded by the viral genome. The 2A protease is highly expressed at an early infection stage and shows high efficiency and selectivity. The cleavage results in loss of FRET and increased CFP signal, which in turn can be detected. The system allowed for detection of 10 PFU or less enteroviruses within 7.5 h postinfection. The reporter cells were shown to be sensitive to coxsackievirus B6, echovirus 11, and poliovirus 1. It was mentioned that the assay might serve as a model for detection of viruses with unique proteases ( Hwang et al. 2006). The assay was successfully tested for detecting infectious echovirus 11 and poliovirus 1 in seeded environmental surface water samples and correlated very well with a plaque assay with the same level of sensitivity (1 pfu). It was also compared with a newly developed IMS + RT-qPCR assay described in the following ( Hwang et al. 2007). However, whereas the plaque test and the IMS + RT-qPCR assay detected all seven human enteroviruses tested, the reporter cell assay detected only the mentioned two enteroviruses with the given reporter cell line. A wider spectrum of enteroviruses might be achieved by using a different reporter cell line with a consensus linker sequence that is recognized by the 2A proteases of all human enteroviruses.
- IMS + RT-qPCR: The assay is based on immunomagnetic separation of enteroviral particles using antibody-coated magnetic beads followed by detection of RNA genomes by molecular beacon-based quantitative RT-PCR. Primers target the 5’ untranslated region. The assay was validated with seeded surface water samples. Seven human enteroviruses (coxsackieviruses B1, B3, B6; Echoviruses 11, 17, 19; and poliovirus 1) could be successfully detected with a sensitivity of 1 pfu. Specificity of enteroviruses was suggested by negative tests for adenovirus 2 and 15 and hepatitis A. A high level of correlation to the conventional plaque assay was found ( Hwang et al. 2007).
- RNA affinity capture + RT-qPCR: Pusch et al. employed affinity capture of enteroviral RNA extracted from environmental samples with Dynabeads. The authors reported that capture efficiency depended strongly on the genomic region chosen for binding. Capture of the RNA by its 3’-end was found to be most efficient. Indirect capture (hybridization of RNA to capture oligonucleotides, then attachment of the resulting duplex molecules to the beads) was found substantially more efficient than direct capture (first attachment of oligonucleotides to the beads, then binding of the RNA). The real-time assay was based on a commercial TaqMan® kit, the primers are given in the original publication ( Pusch et al. 2005).
- NASBA: Rutjes et al. 2005 applied a commercially available quantitative NASBA assay (with internal control RNA) on detection of enterovirus genomes in large volume water samples (10 l sewage and app. 600 l river water). According to the manufacturer (BioMérieux, Boxtel, The Netherlands) the assay detects 263 copies of in vitro transcribed enterovirus RNA per nucleic acid extract with a 95% hit rate using magnetic extraction. End-point NASBA assays for enterovirus detection had already been presented earlier for clinical samples ( Fox et al. 2002; Heim and Schumann 2002; Landry et al. 2003). Enteroviruses were concentrated with nonmagnetic and magnetic NucliSens RNA isolation methods, the latter was found more efficient. In direct comparison, quantitative NASBA was found slightly less sensitive than an in-house RT-qPCR assay although the commercial kit provides high reproducibility as a result of standardization and took only 4 hours. In river samples with virus concentrations below 0.6 PFU per liter, cell culture detection was found more sensitive than either molecular method ( Rutjes et al. 2005).
The following pictures are part of the Public Health Image Library (PHIL) from the Centers for Disease Control and Prevention (CDC). The photos are in the public domain and thus free of copyright restrictions. We would like to express our appreciation for providing these images.
Using immunoelectron microscopic technique, one is able to discern the morphologic traits of the Coxsackie B4 virus virions.
Photo ID: 5630
Content provider(s): Centers for Disease Control and Prevention
This transmission electron micrograph (TEM) revealed the presence of coxsackie B3 virus particles, which were found withing a specimen of muscle tissue. The coxsackie B3 virus is a member of the Picornaviridae family of viruses, and the genus, Enterovirus, as is the well-known, nearly eliminated, Poliovirus.
Photo ID: 10204
Content provider(s): Centers for Disease Control and Prevention/ Dr. Fred Murphy; Sylvia Whitfield