Abstract
Due to the continued persistence of waterborne viral-associated infections, the presence of enteric viruses is a concern. Notwithstanding the health implications, viral diversity and abundance is an indicator of water quality declination in the environment. The aim of this study was to evaluate the presence of viruses (bacteriophage and enteric viruses) in a highly polluted, anthropogenic-influenced river system over a 6-month period at five sampling points. Cytopathic-based tissue culture assays revealed that the isolated viruses were infectious when tested on Hep-G2, HEK293 and Vero cells. While transmission electron microscopy (TEM) revealed that the majority of the viruses were bacteriophages, a number of presumptive enteric virus families were visualized, some of which include Picornaviridae, Adenoviridae, Polyomaviridae and Reoviridae. Finally, primer specific nested polymerase chain reaction (nested-PCR)/reverse transcription-polymerase chain reaction (RT-PCR) coupled with BLAST analysis identified human adenovirus, polyomavirus and hepatitis A and C virus genomes in river water samples. Taken together, the complexity of both bacteriophage and enteric virus populations in the river has potential health implications. Finally, a systematic integrated risk assessment and management plan to identify and minimize sources of faecal contamination is the most effective way of ensuring water safety and should be established in all future guidelines.
INTRODUCTION
The contamination of water bodies with diverse microbial communities is a major concern to public health (Pandey et al. 2014). Diarrhoeal outbreaks caused by waterborne pathogens is alarming, with most cases reported in developing countries due to the lack of proper drinking water and sanitation infrastructure (Hofstra 2011). The World Health Organization (WHO) approximates 1.8 billion people worldwide currently utilizing drinking water resources that are contaminated by faecal pollution (WHO 2017). Among these pathogens, the greatest cause for concern comes from the prevalence of enteric viruses in water (Leifels et al. 2016). Enteric viruses are excreted from the faeces of infected individuals at very high numbers (1011/g faeces) while concomitantly exceeding other existing microbial communities by 10–100-fold in water environments (Ganesh & Lin 2013). Apart from the socio-economic implications of waterborne virus-related sicknesses in both developing and developed nations, the extent of the burden and impact of viral disease is far concentrated in regions with enormous environmental contaminations (Rodríguez-Díaz et al. 2009). Over 140 different types of viruses are known to infect humans, resulting in a variety of illnesses (Chigor & Okoh 2012). While gastroenteritis is the most common outcome, several acute illnesses such as meningitis, hepatitis, conjunctivitis and muscular syndromes (myocarditis, fibromyalgia) are associated with enteric viral infections. Waterborne viral infections have also been implicated in some chronic diseases such as chronic fatigue syndrome and diabetes (La Rosa et al. 2012). Enteric viruses enter water systems through agricultural and urban runoffs, septic systems, sewage outfalls and wastewater discharge which are then transmitted through the faecal–oral route to finally replicate in the gastrointestinal tract of humans and animals (Coudray-Meunier et al. 2013).
Among the many types of enteric viruses reported in aquatic environments, the most common include hepatitis A viruses (Dongdem et al. 2009), adenoviruses (Chigor & Okoh 2012), polyomaviruses (Haramoto et al. 2010) and noroviruses (Lopman et al. 2016). Hepatitis A viruses cause approximately 40% of severe hepatitis annually (Redwan & Abdullah 2012) with numerous outbreaks reported on a global scale (Frank et al. 2007; Robesyna et al. 2009; Chen et al. 2017). Furthermore, in comparison to hepatitis B, infection with hepatitis A viruses are three-fold greater in travellers (Redwan & Abdullah 2012). Although the incidence of hepatitis A infections varies across countries, sporadic hepatitis A outbreaks are mostly observed in non-industrialized regions with poor hygiene conditions (La Rosa et al. 2014). Adenoviruses cause a variety of clinical symptoms with many outbreaks related to recreational water exposure (Vieira et al. 2012). Adenovirus detection is simple and their numbers far exceed hepatitis A and enteroviruses in many aquatic habitats. Furthermore, adenoviruses show greater stability to chlorination and UV irradiation than enteroviruses (Hundesa et al. 2006), allowing them to persist in the environment for much longer time periods. Human BK and JC polyomaviruses cause chronic infections in humans and are frequently defecated in municipal sewage and urine. These viruses have been linked to other important diseases in immune-compromised individuals as well as some types of human cancer such as colorectal cancer (Hundesa et al. 2006). The ubiquitous noroviruses are the most common cause of gastroenteritis, associated with 18% of global diarrhoea cases and responsible for 212,000 deaths yearly on a global scale (Lopman et al. 2016). Even though hepatitis C viruses are not waterborne but fluid related (blood-borne), they are important aetiological agents of non-A, non-B hepatitis in people (Aslanzadeh et al. 1996). East and Central Asia as well as northern Africa are regions mostly affected by hepatitis C infections (WHO 2016). Beld et al. (2000) frequently detected hepatitis C viruses in the faeces of chronically infected patients. Therefore, speculation of their possible presence in water through surface runoffs could be made. Moreover, studies involving the transmission of hepatitis C viruses in injection drug users have demonstrated the presence of these viruses in water containers (Doerrbecker et al. 2013) and rinse water (Thorpe et al. 2002).
Current water quality guidelines mostly rely on bacterial indicators of faecal pollution. However, bacterial pathogens have been poorly correlated with the presence of enteric viruses (Lee et al. 2014) and are more susceptible than viruses to some water treatment processes (Rodríguez et al. 2008). Bacteriophages (or phages) are viruses that infect bacteria and are recommended as alternate indicators of faecal pollution (Ganesh et al. 2014). Apart from morphological and internal chemical similarities to enteric viruses, bacteriophage detection is usually cost-effective and easier to apply than enteric virus detection (Leclerc et al. 2000). However, the recommended plaque assays require long experimental time frames, are dependent on bacterial host viability and suffer from errors during enumeration (Edelman & Barletta 2003). Importantly, Wu et al. (2011) demonstrated via a logistic regression analysis that the correlation between coliphages and pathogens in water was poor (p = 0.186) where only 40 out of the 85 cases were correlated.
To investigate the presence and diversity of viruses (bacteriophages and enteric viruses) in the Umhlangane River, a tangential flow filtration (TFF) process was adopted to concentrate the virus particles. This method has been reported to effectively concentrate virus particles by reducing filter clogging via parallel fluid movement across the membrane filters. Moreover, TFF ensures that pore sizes and cross-flow prevents bacterial contamination of the viral concentrate (Cai et al. 2015). Together with TFF, various comparative detection methods including transmission electron microscopy (TEM) and tissue culture were used to assess the nature of the virus particles. Finally, a nested polymerase chain reaction (nested-PCR)/reverse transcription-polymerase chain reaction (RT-PCR) was used to evaluate the presence of human adenovirus, polyomavirus and hepatitis A and C virus genomes in the Umhlangane River.
METHODS
Water sample collection
The Umhlangane River, located in the heart of Durban is surrounded by a plethora of both developed and undeveloped societies. The river spans approximately 15 km in length and drains into KwaZulu-Natal's main drinking water catchments, the Umgeni River (Hadlow 2011). In addition to being a conduit for domestic and industrial wastes, the Umhlangane River receives animal waste from nearby farms. Moreover, the surrounding informal settlement or squatter camp communities directly utilize the river water for drinking as well as other domestic practices (Hadlow 2011).
Twenty litres (20 L) of river water was collected at five different sampling points (designated P1 to P5), described in Table 1. Each sampling site was selected based on different land use zones to assess the extent of its influence on the catchment. Sampling was conducted in the second week of every month over a 6-month period, commencing in April 2014 and concluding in September 2014, for virus concentration and analysis. Water samples were collected in 25 L plastic drums (previously disinfected with 70% (v/v) alcohol and rinsed with deionized water). At each sampling point, the drums were rinsed with river water prior to being plunged approximately 0.3–0.5 m below the water surface to circumvent the disinfectant effect of UV light (Jurzik et al. 2010). Samples were transported on ice to the Discipline of Microbiology, University of KwaZulu-Natal (Westville campus) for virus concentration.
Coordinates and description of the five sampling points
Sampling points . | GPS coordinates . | Site description . | |
---|---|---|---|
Latitude . | Longitude . | ||
P1 | 29° 42′47″S | 30° 59′33″E | Phoenix industrial |
P2 | 29° 43′35″S | 31° 00′21″E | Upstream KwaMashu wastewater treatment plant |
P3 | 29° 43′35″S | 30° 00′21″E | Natural wetlands |
P4 | 29° 45′39″S | 30° 01′11″E | Riverhorse Valley business estate |
P5 | 29° 46′10″S | 30° 00′24″E | Springfield industrial |
Sampling points . | GPS coordinates . | Site description . | |
---|---|---|---|
Latitude . | Longitude . | ||
P1 | 29° 42′47″S | 30° 59′33″E | Phoenix industrial |
P2 | 29° 43′35″S | 31° 00′21″E | Upstream KwaMashu wastewater treatment plant |
P3 | 29° 43′35″S | 30° 00′21″E | Natural wetlands |
P4 | 29° 45′39″S | 30° 01′11″E | Riverhorse Valley business estate |
P5 | 29° 46′10″S | 30° 00′24″E | Springfield industrial |
Primary virus concentration
A modified TFF process described by Ganesh et al. (2014) was used to concentrate viruses from the collected water samples (Figure 1). Briefly, 20 L of river water was passed through a 0.45 μm sediment filter (Merck-Millipore Corp.) at 130 mL/min to remove large debris and solids. Virus concentration then involved two separate steps. First, the water was filtered through a 142 mm diameter, 0.22 μm membrane filter (Merck-Millipore Corp.) at 330 mL/min to remove all bacteria. The second step further concentrated the viruses from the 0.22 micron concentrate through a 100 kDa (molecular weight cut-off) cartridge filter. The resulting retentate was then re-circulated through the system until approximately 500 mL of sample remained.
Schematic diagram of TFF procedure to concentrate viral particles from the river water samples (adapted from Ganesh et al. (2014) with some modifications). Debris removal and bacterial concentrate (a), removal of bacteria (b), feed (c), retentate (d) and permeate (e).
Schematic diagram of TFF procedure to concentrate viral particles from the river water samples (adapted from Ganesh et al. (2014) with some modifications). Debris removal and bacterial concentrate (a), removal of bacteria (b), feed (c), retentate (d) and permeate (e).
Secondary virus concentration
Re-concentration of the TFF samples was carried out using ultracentrifugation according to a procedure described by Colombet et al. (2007) with some modifications. Six tubes of 28 mL retentate were ultracentrifuged for 2.5 hours at 130,000 × g (i.e., 29,000 revolutions per minute (rpm); 4 °C) in a SW-32 Ti rotor (Optima L-100 XP, Beckman Coulter). The viral pellets were re-suspended in 500 μL phosphate buffered saline (PBS; pH 7.2), pooled together (final volume of 3 mL) and stored at −20 °C until further analyses.
Tissue culture experiments
Prior to presumptive identification, cell lines were used to determine the infectious nature of the viruses in the river water. Viruses concentrated on the first (April) and fourth (July) months were tested for infectivity. Three different cell lines were used: (i) human hepatocellular carcinoma (Hep-G2), (ii) African green monkey kidney cells (Vero) and (iii) human embryonic kidney cells (HEK293). The cells were monitored daily using an inverted microscope (Olympus) at 400 × magnification for the production of a cytopathic effect (CPE) indicating positive virus infectivity. Adenovirus 9 VR-1086 and hepatitis A VR-1402 controls were tested on the three cell lines during both experiments. Detailed experimental procedures are described in Ganesh et al. (2014).
TEM for viral diversity
Viral morphology and diversity was examined using TEM. After secondary concentration, one drop of each water sample was placed onto a carbon-coated grid (Electron Microscopy Sciences, Fort Washington, PA) for 2 minutes, stained with 4% uranyl acetate solution for 30 s, rinsed with deionized water for 10 s and air dried prior to visualization with a high resolution TEM (JEOL 2100). Electron micrographs of the virus particles were taken between 250,000 and 500,000 × magnification. Bacteriophage and virus particles were measured for size and compared to known viruses previously described in the literature for presumptive identification.
Molecular detection of human enteric viruses in the Umhlangane River
Viral nucleic acids (total RNA and DNA) were extracted separately from 1 mL samples each using the High Pure Viral RNA and DNA Kits (Roche Diagnostics, Germany), respectively. The quantity and quality of the extracts were measured using the NanoDrop 2200 spectrophotometer (Thermo Scientific, Finland). For RNA samples, first strand cDNA synthesis was carried out using the DyNamo™ cDNA Synthesis Kit (Finnzymes, Thermo Fischer Scientific, Finland). Thereafter, primer specific (Table 2) nested-PCR was performed for four viral populations as described below. Five microlitres of the PCR product from the first round was used as a template for the second round for all reactions.
Primer sequences for the PCR amplification of the four viral groups
Primer . | Primer sequence (5′ − 3′) . | Amplicon size . | Reference . |
---|---|---|---|
Adenoviruses (A-F) | |||
AV-A1 | GCCGCAGTGGTCTTACATGCACATC | 300 bp | Allard et al. (1992) |
AV-A2 | CAGCACGCCGCGGATGTCAAAGT | ||
AV-B1a | GCCACCGAGACGTACTTCAGCCTG | 143 bp | |
AV-B2a | TTGTACGAGTACGCGGTATCCTCGCGGTC | ||
Polyomaviruses | |||
P1 | GTATACACAGCAAAGGAAGC | 630 bp | McQuaig et al. (2006) |
P2 | GCTCATCAGCCTGATTTTGG | ||
P3a | AGTCTTTAGGGTCTTCTACC | 173 bp | |
P4a | GGTGCCAACCTATGGAACAG | ||
Hepatitis A viruses | |||
HHA1 | TGCAAATTAYAAYCAYTCTGATGA | 532 bp | Pina et al. (2001) |
HHA2 | TTTCTGTCCATTTYTCATCATTC | ||
HHA3a | TTYAGTTGYTAYTTGTCTGT | 436 bp | |
HHA4a | TCAAGAGTCCACACACTTC | ||
Hepatitis C viruses | |||
HCV1 | ACTGTCTTCACGCAGAAAGCGTCTAGCCAT | 271 bp | Hu et al. (2003) |
HCV2 | CGAGACCTCCCGGGGCACTCGCAAGCACCC | ||
HCV3a | ACGCAGAAAGCGTCTAGCCATGGCGTTAGT | 255 bp | |
HCV4a | TCCCGGGGCACTCGCAAGCACCCTATCAGG |
Primer . | Primer sequence (5′ − 3′) . | Amplicon size . | Reference . |
---|---|---|---|
Adenoviruses (A-F) | |||
AV-A1 | GCCGCAGTGGTCTTACATGCACATC | 300 bp | Allard et al. (1992) |
AV-A2 | CAGCACGCCGCGGATGTCAAAGT | ||
AV-B1a | GCCACCGAGACGTACTTCAGCCTG | 143 bp | |
AV-B2a | TTGTACGAGTACGCGGTATCCTCGCGGTC | ||
Polyomaviruses | |||
P1 | GTATACACAGCAAAGGAAGC | 630 bp | McQuaig et al. (2006) |
P2 | GCTCATCAGCCTGATTTTGG | ||
P3a | AGTCTTTAGGGTCTTCTACC | 173 bp | |
P4a | GGTGCCAACCTATGGAACAG | ||
Hepatitis A viruses | |||
HHA1 | TGCAAATTAYAAYCAYTCTGATGA | 532 bp | Pina et al. (2001) |
HHA2 | TTTCTGTCCATTTYTCATCATTC | ||
HHA3a | TTYAGTTGYTAYTTGTCTGT | 436 bp | |
HHA4a | TCAAGAGTCCACACACTTC | ||
Hepatitis C viruses | |||
HCV1 | ACTGTCTTCACGCAGAAAGCGTCTAGCCAT | 271 bp | Hu et al. (2003) |
HCV2 | CGAGACCTCCCGGGGCACTCGCAAGCACCC | ||
HCV3a | ACGCAGAAAGCGTCTAGCCATGGCGTTAGT | 255 bp | |
HCV4a | TCCCGGGGCACTCGCAAGCACCCTATCAGG |
bp, base pair.
aNested primers.
Following nested-PCR, random positive bands were selected with their primer sets and sequenced (Inqaba Biotech, South Africa). The sequences were analysed by BLAST (http://www.ncbi.nlm.nih.gov/BLAST) to confirm the identities of the presumptive positive PCR products.
Human adenoviruses
The hexon gene of 47 different adenovirus serotype genomes were amplified (Allard et al. 1992). Both rounds contained an additional 0.4 mM MgCl2. The PCR conditions were as follows: 4 min at 94 °C, 40 cycles of 92 °C for 30 s, 60 °C for 30 s and 72 °C for 1 min. Final elongation was at 72 °C for 5 min. A positive control (cell cultured adenovirus) was included for all reactions.
Human polyomaviruses
The human-specific BK and JC polyomavirus genomes were amplified according to McQuaig et al. (2006) with some modifications. Amplification conditions were carried out accordingly: 94 °C for 2 min, 45 cycles of 94 °C for 20 s, 55 °C for 20 s, 72 °C for 20 s. Final elongation was 72 °C for 2 min. Both rounds had an additional 0.5 mM MgCl2.
Hepatitis A viruses
The VP1/VP2 region of the hepatitis A virus genomes were amplified (Pina et al. 2001). The PCR conditions included: 95 °C for 3 min, 30 cycles of 95 °C for 60 s, 42 °C for 60 s and 72 °C for 60 s. Final extension was 72 °C for 5 min. Both rounds contained an additional 0.5 mM MgCl2.
Hepatitis C viruses
The 5′ untranslated region (5′ UTR) of the hepatitis C virus genomes were amplified according to Hu et al. (2003) with some modifications. The PCR were run under the following conditions: 95 °C for 3 min, 30 cycles of 95 °C for 60 s, 50 °C for 60 s and 72 °C for 60 s. Final extension was 72 °C for 5 min.
Quality control
The probability of sample contamination due to DNA amplicons or cross-contamination was reduced through the practice of standard molecular preparation protocols. Separate areas were used to prepare the reagents and manipulate the amplified samples. Negative controls were included in all reactions and a positive control was also used where available. All RNA samples were manipulated in a separate RNA room (DNase/RNase-free zone) that contained PCR pipettes, filter tips, centrifuges, etc., specifically designated for RNA work only. Furthermore, master-mixes for the PCR reactions were prepared in the bio-safety cabinet to prevent contamination.
RESULTS
Virus infectivity
The CPE of the viral concentrate was based on morphological changes of the cells, the visibility of granulated, elongated cells, vacuole production and the loss of cell-to-cell adherence and the wall of the flask. The viral concentrate was capable of causing infectivity to all three cell lines with the exception of P3 (Vero) and P4 (HEK293; Vero) in April as well as P4 (Hep-G2) and P5 (Hep-G2; Vero) in July. Interestingly, while some morphological changes were much more apparent than others, some cells produced a much later CPE. In addition to clearer, more apparent morphological changes, the control-infected cells produced a CPE quicker than the concentrated viruses. However, in comparison to the adenovirus 9 VR-1086 control which was CPE positive on all cell lines, the hepatitis A VR-1402 control was CPE positive on the Hep-G2 and Vero cells only during both experiments.
Visualization and presumptive identification of bacteriophage populations
TEM revealed a number of phage morphotypes at all sampling locations which are depicted in Figure 2. Bacteriophages were identified according to the International Committee on the Taxonomy of Viruses (ICTV) classification scheme described in Ackermann & Eisenstark (1974). Bacteriophages belonging to the Caudovirales order consisting of morphotypes A1 (Myoviridae – contractile tails), B1 (Siphoviridae – short capsid, non-contractile and long tail), B2 (Siphoviridae – long capsid, non-contractile and long tail) and C1 (Podoviridae – short tail) were identified in the Umhlangane River. Siphoviridae members B1 and B2 closely resembled the phages seen in Figure 2(a) and 2(g)), respectively while Figure 2(c) and 2(d)) closely resemble members of the Podoviridae and Myoviridae families.
TEM images of diverse phage morphotypes visualized in the Umhlangane River (a)–(h). Bacteriophages (i), (k), (m) resemble the (j) known T4-like Vibrio parahaemolyticus phage (Ackermann & Heldal 2010), (l) known Mycobacterium 40AC phage (Stella et al. 2013) and known (n) VvAWI Vibrio vulnificus phage (Nigro et al. 2012). Images captured at 250,000 to 500,000 × magnification. Scale bar 100 nm.
TEM images of diverse phage morphotypes visualized in the Umhlangane River (a)–(h). Bacteriophages (i), (k), (m) resemble the (j) known T4-like Vibrio parahaemolyticus phage (Ackermann & Heldal 2010), (l) known Mycobacterium 40AC phage (Stella et al. 2013) and known (n) VvAWI Vibrio vulnificus phage (Nigro et al. 2012). Images captured at 250,000 to 500,000 × magnification. Scale bar 100 nm.
While most bacteriophages comprised regular hexagonal outlined heads (Figure 2(e) and 2(f)), some were irregular (Figure 2(b)). The discrimination between octahedral, icosahedral and dodecahedral shapes could not be properly determined. Furthermore, the tail fibres, neck and baseplate were only visualized in Figure 2(d) (tail fibres and neck) and Figure 2(e) (neck and baseplate). Myoviridae tail contraction was observed in Figure 2(h).
Figure 2(i)–2(n)) represent phages identified in the Umhlangane River at the five sampling points that resemble some known bacteriophages. These phages include the T4-like Vibrio parahaemolyticus phage (Ackermann & Heldal 2010), the environmentally isolated Mycobacterium 40AC phage (Stella et al. 2013) and the VvAWI Vibrio vulnificus phage (Nigro et al. 2012).
TEM for presumptive enteric virus identification
Figures 3 and 4 illustrate presumptive enteric viruses that were observed during the sampling period. Classification was performed according to size measurements and comparative structural similarities to known viruses found in the literature. Presumptive naked enterovirus-like particles (Picornaviridae) with sizes ranging from 25.92 to 27.46 nm (Figure 3(a)–3(c)) were visualized and compared to known coxsackieviruses (Figure 3(d)). Figure 3(e)–3(g)) depict the TEM images of presumptive naked Adenoviridae-like particles and (Figure 3(h)) known adenoviruses (70–90 nm). The viral particle sizes ranged from 67.29 to 78.11 nm. Although less commonly observed, TEM also revealed presumptive Polyomaviridae-like particles (Figure 3(i) and 3(j)), Reoviridae-like particles (Figure 3(l)–3(n)) and Coronaviridae-like particles (Figure 4(a)–4(c)) when compared against known polyomaviruses (40–50 nm; Figure 3(k)), known rotaviruses (Figure 3(o)) and a known coronavirus (Figure 4(d)). Herpesviridae-like particles (Figure 4(e)–4(g)) and Orthomyxoviridae-like particles (Figure 4(i)–4(k)) were compared to known herpes (Figure 4(h)) and influenza viruses (Figure 4(l)), respectively. Lastly, many presumptive enveloped viruses (Figure 4(m)–4(p)) were seen in the Umhlangane River. However, due to inherently similar structures, comparisons between enveloped viruses and known viruses could not be made. Interestingly, TEM revealed some important morphological characteristics, such as the double-layered rotavirus (Figure 3(n)), the Coronaviridae peplomers or ‘setting-sun’ appearance (Figure 4(a)), the rough polyomavirus capsids (Figure 3(j)) and the envelopes and nucleocapsids of many viruses (Figure 4(f), 4(j) and 4(m)).
TEM images of presumptive enteric viruses. Picornaviridae-like particles (a)–(c), (d) known coxsackievirus (Schramlová et al. 2010), Adenoviridae-like particles (e)–(g), (h) known adenoviruses (Li et al. 2013), Polyomaviridae-like particles (i) and (j), (k) known polyomaviruses (Broekema & Imperiale 2012), Reoviridae-like particles (l)–(n) and (o) known rotaviruses (Zeng et al. 1996). Images captured at 250,000 to 500,000 × magnification. Scale bar 100 nm.
TEM images of presumptive enteric viruses. Picornaviridae-like particles (a)–(c), (d) known coxsackievirus (Schramlová et al. 2010), Adenoviridae-like particles (e)–(g), (h) known adenoviruses (Li et al. 2013), Polyomaviridae-like particles (i) and (j), (k) known polyomaviruses (Broekema & Imperiale 2012), Reoviridae-like particles (l)–(n) and (o) known rotaviruses (Zeng et al. 1996). Images captured at 250,000 to 500,000 × magnification. Scale bar 100 nm.
TEM images of presumptive enteric viruses. Coronaviridae-like particles (a)–(c), (d) known coronavirus (Schramlová et al. 2010), Herpesviridae-like particles (e)–(g), (h) known herpes viruses (Goldsmith & Miller 2009), Orthomyxoviridae-like particles (i)–(k), (l) known influenza viruses (Schramlová et al. 2010) and enveloped virus-like particles (m)–(p). Images captured at 250,000 to 500,000 × magnification. Scale bar 100 nm.
TEM images of presumptive enteric viruses. Coronaviridae-like particles (a)–(c), (d) known coronavirus (Schramlová et al. 2010), Herpesviridae-like particles (e)–(g), (h) known herpes viruses (Goldsmith & Miller 2009), Orthomyxoviridae-like particles (i)–(k), (l) known influenza viruses (Schramlová et al. 2010) and enveloped virus-like particles (m)–(p). Images captured at 250,000 to 500,000 × magnification. Scale bar 100 nm.
Molecular detection of four enteric viral groups
Human adenovirus, polyomavirus, hepatitis A and C virus genomes detected using a nested PCR/RT-PCR are depicted in Table 3. All sampling points during the 6-month sampling period produced a positive PCR product of 143 bp corresponding to the adenovirus control. Positive BK and JC human polyomavirus genomes were detected in 60% (n = 30) of the tested samples, yielding the 173 bp expected product size. All samples tested during June were positive for polyomaviruses while only P1 (Phoenix industrial) during September depicted a positive PCR product (Table 3). The VP1/VP2 regions of hepatitis A viruses were detected in 70% (n = 30) of the samples yielding 436 bp products. The expected product size of 255 bp amplifying the 5′ UTR for positive hepatitis C viruses was only seen in three samples, including P2 in May (upstream KwaMashu WWT), P4 in June (Riverhorse Valley business estate) and P5 in July (Springfield industrial). Finally, correlations between seasonality and the presence of the viral groups were not observed.
Nested-PCR/RT-PCR amplification of the four viral groups
Viruses . | Month . | Location . | ||||
---|---|---|---|---|---|---|
P1 . | P2 . | P3 . | P4 . | P5 . | ||
Adenovirus | April | + | + | + | + | + |
May | + | + | + | + | + | |
June | + | + | + | + | + | |
July | + | + | + | + | + | |
August | + | + | + | + | + | |
September | + | + | + | + | + | |
Polyomavirus | April | + | + | − | + | − |
May | + | + | + | − | − | |
June | + | + | + | + | + | |
July | + | − | − | + | − | |
August | + | − | + | + | + | |
September | + | − | − | − | − | |
Hepatitis A virus | April | − | + | + | + | + |
May | − | + | − | − | + | |
June | + | + | − | + | + | |
July | − | + | + | + | + | |
August | + | + | + | + | + | |
September | + | − | + | − | − | |
Hepatitis C virus | April | − | − | − | − | − |
May | − | + | − | − | − | |
June | − | − | − | + | − | |
July | − | − | − | − | + | |
August | − | − | − | − | − | |
September | − | − | − | − | − |
Viruses . | Month . | Location . | ||||
---|---|---|---|---|---|---|
P1 . | P2 . | P3 . | P4 . | P5 . | ||
Adenovirus | April | + | + | + | + | + |
May | + | + | + | + | + | |
June | + | + | + | + | + | |
July | + | + | + | + | + | |
August | + | + | + | + | + | |
September | + | + | + | + | + | |
Polyomavirus | April | + | + | − | + | − |
May | + | + | + | − | − | |
June | + | + | + | + | + | |
July | + | − | − | + | − | |
August | + | − | + | + | + | |
September | + | − | − | − | − | |
Hepatitis A virus | April | − | + | + | + | + |
May | − | + | − | − | + | |
June | + | + | − | + | + | |
July | − | + | + | + | + | |
August | + | + | + | + | + | |
September | + | − | + | − | − | |
Hepatitis C virus | April | − | − | − | − | − |
May | − | + | − | − | − | |
June | − | − | − | + | − | |
July | − | − | − | − | + | |
August | − | − | − | − | − | |
September | − | − | − | − | − |
+, positive PCR product; −, negative PCR product. Note: italicized font, autumn; normal font, winter; underlined font, spring.
Comparisons between random positive PCR products and reference sequences in GenBank confirmed the presence of adenovirus, polyomavirus and hepatitis A and C virus genomes in the Umhlangane River. The percentage maximum identity and E-values obtained during the BLAST analysis are depicted in Table 4. The sequence identities ranged between 86% and 100% to their known complements on the GenBank database. Among the sequenced data, human adenovirus C strain, JC polyomavirus isolate GCN8, hepatitis A virus strain CFH and hepatitis C virus isolate Ind-MN19 were identified.
BLAST analysis showing the maximum identity and E-values for the tested samples
Accession . | Description . | E-value . | Maximum identity (%) . |
---|---|---|---|
KF268310.1 | Human adenovirus c strain | 2e-28 | 89 |
KM205587.1 | Adenovirus 2 isolate AAU4 | 1e-26 | 86 |
KM225765.1 | JC polyomavirus isolate GCN8, complete genome | 9e-53 | 98 |
AB081021.1 | JC virus DNA, isolate ME-5 | 4e-61 | 99 |
HQ246217.1 | Hepatitis A virus CFH-HAV, complete genome | 3e-177 | 95 |
FJ687513.1 | Hepatitis A isolate 9 polyprotein | 4e-166 | 96 |
EF473252.1 | Hepatitis C isolate Ind-MN19 5′ UTR | 8e-05 | 100 |
Accession . | Description . | E-value . | Maximum identity (%) . |
---|---|---|---|
KF268310.1 | Human adenovirus c strain | 2e-28 | 89 |
KM205587.1 | Adenovirus 2 isolate AAU4 | 1e-26 | 86 |
KM225765.1 | JC polyomavirus isolate GCN8, complete genome | 9e-53 | 98 |
AB081021.1 | JC virus DNA, isolate ME-5 | 4e-61 | 99 |
HQ246217.1 | Hepatitis A virus CFH-HAV, complete genome | 3e-177 | 95 |
FJ687513.1 | Hepatitis A isolate 9 polyprotein | 4e-166 | 96 |
EF473252.1 | Hepatitis C isolate Ind-MN19 5′ UTR | 8e-05 | 100 |
DISCUSSION
Extensive concentration methods and/or the analysis of large volumes of water are usually required to obtain a sufficient amount of viruses for experimental testing (Symonds & Breitbart 2015). In this study, TFF coupled with ultracentrifugation efficiently concentrated viruses from the collected water samples. In addition to clear, bacterial-free concentrates, this two-step TFF procedure ensured that most viruses remained structurally intact.
The detection of infectious enteric viruses in water has long employed cell culture methods (Calgua et al. 2011). Previously, this was the only method approved by the Environmental Protection Agency (EPA) for the detection of enteric viruses (Jiang 2006). In 2012, the EPA modified Method 1615 to incorporate culture and quantitative RT-PCR (RT-qPCR) for the detection of norovirus and enterovirus (Cashdollar et al. 2013). Positive CPE produced by the viral concentrates indicates the presence of infectious enteric viruses in the Umhlangane River. Sampling points P1 (Phoenix industrial) and P2 (upstream KwaMashu WWTP) depicted positive CPE on all cell lines. This may be due to the influx of effluents containing human or animal faecal matter harbouring some enteric viruses (La Rosa et al. 2012). Cytopathic-based tissue culture assays (or quantal assays) are sensitive rather than quantitative (EPA 2001). Quantal assays are simplistic since the effect is either present or not (Zivin & Waud 1992) and in tissue culture only one infectious particle is enough to produce a successful CPE (EPA 2001). However, the efficacy of viral replication on various cell lines depends on the serotypes of the viruses present (Jiang 2006). Therefore, not all viruses or serotypes are susceptible to all cell lines (Lee et al. 2004). Furthermore, some enteric viruses are slow growing (Jiang et al. 2009) or produce unclear or no CPE (Calgua et al. 2011). Consequently, the negative CPE produced by some of the samples may be due to viral diversity or their slow growing nature, indicating negative, little or no CPE on the cell monolayers. Moreover, diverse viral concentrates can inhibit the growth of some viruses due to the interference by other groups of viruses (Carducci et al. 2002). In the present study, Hep-G2, HEK293 and Vero cell lines were used. The Hep-G2 and HEK293 carcinoma cell lines are common and can support the growth of many viruses (Leland & Ginocchio 2007). In particular, Hep-G2 cells are highly sensitive to hepatitis A, B and C viruses (WHO 2008) while HEK293 cells show great sensitivity to human adenovirus types 40 and 41 (Jiang et al. 2009). Since cell lines have an affinity to propagate certain viral groups the Vero cell line was used to increase the sensitivity of viruses in this study. Vero cells can support the growth of hepatitis viruses (Konduru & Kaplan 2006), measles viruses, rubella viruses and arboviruses (Osada et al. 2014).
TEM was used to visualize the types of viruses present in the Umhlangane River. Bacteriophages belonging to the Myoviridae, Siphoviridae and Podoviridae families were morphologically diverse showing great abundance. Bacteriophage diversity corresponds with the colossal dynamics and diversity observed in their bacterial counterparts and vice versa (Demuth et al. 1993; Beckett & Williams 2013; Williams 2013). The complexity between phage and bacterial populations were reiterated by the presumptive identification of phages described to those confirmed in the literature (Figure 2(i)–2(n)). For example, non-tuberculosis-causing, environmental Mycobacteria may be found in water and soil (Pedulla et al. 2003). However, bacteriophages could originate from allochthonous sources (Hewson et al. 2012), such as wild animals, waterfowl or anthropogenic practices near the river (Demuth et al. 1993).
A number of diverse enteric viruses observed in the Umhlangane River indicates a significant amount of faecal contamination is entering the catchment (Cantalupo et al. 2011). The source of pollution and consequent viral contamination may be from surface runoff, recreational activities, storm water discharge, sewage discharge and overflows as well as other humanized practices (Olaniran et al. 2009). Many of the detected viruses were found upstream in P1 and P2 (Phoenix residential/industrial; upstream KwaMashu WWTP) and downstream in P4 and P5 (Riverhorse Valley business estate; Springfield industrial). These areas are densely populated, comprising a variety of anthropogenic practices with the additional input of animal faecal matter from the surrounding farms (Phoenix areas) and effluent from informal settlements (Riverhorse Valley). This may have contributed to the viruses observed in the river water samples.
Presumptive Polyomaviridae, Coronaviridae, Adenoviridae, Picornaviridae, Herpesviridae, Reoviridae (rotavirus), Orthomyxoviridae and other unassigned enveloped viruses were visualized. Viruses with intact nucleocapsids are protected by an envelope while naked viruses comprise inflexible capsids that can withstand harsh conditions (Ackermann & Heldal 2010). These morphological characteristics allow enteric viruses to persist in the environment for long periods of time and, most often than not, survive many treatment processes (Steyer et al. 2011). However, since these viruses were concentrated from water, their morphological structures may have faulted through degradation effects (Eifan 2013) or via the concentration method (Karim et al. 2004). Finally, morphological characteristics coupled with seasonal periods and environmental factors are important points to consider when assessing viral persistence (Espinosa et al. 2009).
The visualization of a diverse array of enteric viruses in the Umhlangane River drove the motivation for the molecular detection of some enteric viruses that may be present in the river water samples. To the best of our knowledge, this is the first study conducted to evaluate the presence of enteric viruses in the Umhlangane River of Durban, South Africa. The application of PCR-based methods to study the presence of enteric viruses in aquatic environments has advanced over the years (Jiang 2006). Although conventional PCR has proved to be effective, specific and sensitive to detect low virus concentrations (Girones et al. 2010), nested-PCR/RT-PCR can further increase the sensitivity of detection by employing two rounds of PCR using two sets of primers (van Heerden et al. 2005). Nested-PCR/RT-PCR was used to detect the presence of human adenovirus, polyomavirus and hepatitis A and C virus populations in this study. First round PCR/RT-PCR was not able to amplify the viral genomes for some (adenovirus), many (polyomavirus) or all of the tested samples (hepatitis A and C viruses). However, the second round of PCR/RT-PCR using the inner primers and primer templates from the first rounds was sensitive enough to detect the viral genomes.
Human adenoviruses were observed in 100% (n = 30) of the tested water samples (Table 3). Currently, there are over 60 types of adenoviruses present within the human adenovirus A–G species (Robinson et al. 2013). Serotypes 40 and 41 are the second leading cause of gastroenteritis in children next to rotaviruses (Jiang et al. 2001).
Human polyomaviruses (BK and JC) were identified in 18 (n = 30) of river water samples (Table 3). The BK and JC polyomaviruses are exclusive to humans and cause asymptomatic viruria (Polo et al. 2004). Although these viruses can be excreted from human faeces (Hundesa et al. 2006), the abundance of human polyomaviruses may arise from urine since JC viruses have been found in 20–80% of adult urine samples (Kitamura et al. 1990). Moreover, these viruses are protected when ingested by food particles and are stable at acidic pH (Bofill-Mas et al. 2001).
The hepatitis A virus is the number one cause for gastroenteritis worldwide (Redwan & Abdullah 2012). These viruses have been successfully isolated from various water sources including dams (Taylor et al. 2001), rivers (Pina et al. 2001) and groundwater (Borchardt et al. 2003). Not only can hepatitis A viruses persist in groundwater for months (La Rosa et al. 2012), they also show resistance to common disinfectants (Bigliardi & Sansebastiano 2006). Furthermore, these viruses can survive exposure to 20% ether, acidity (pH 1.0 for 2 hours) and heating to 60 °C for 1 hour (Kocwa-Haluch 2001).
Although hepatitis C viruses were only detected in 10% (n = 30) of the collected water samples (Table 3), their presence is a cause for concern. These viruses infect approximately 150 to 200 million people worldwide (Maier & Wu 2002). Hepatitis C viruses cause chronic and acute hepatic diseases and, due to high viral mutation rates, no licensed therapeutic or prophylactic vaccines have been made available (Abdelwahab & Said 2016).
Although this study could not identify patterns of viral seasonal occurrence (Table 3), it is important to note that some viruses, such as rotaviruses, hepatitis A (Chigor & Okoh 2012) and astroviruses (Ganesh & Lin 2013), to name a few, are affected by seasonal conditions.
The nucleotide sequences of the adenovirus, polyomavirus and hepatitis A and C virus genomes confirmed using BLAST analysis revealed different virus strains. Not only does the BLAST analysis confirm the presence of these four viral groups, it also indicates the diversity of these enteric viruses residing in the Umhlangane River.
CONCLUSIONS
In conclusion, this study investigated the presence of viral populations in the Umhlangane River using TFF and subsequent molecular methods. The TFF concentration of viruses from 20 L of river water was effective and did not require sample manipulation such as pH adjustment or PCR inhibitors (beef extract) for elution (Jiang et al. 2001). When assessed for infectivity, the viral concentrates were found to be infectious to Hep-G2, HEK293 and Vero cell lines. Furthermore, the use of TEM allowed the visualization of diverse bacterial and eukaryotic viruses present in the catchment. Molecular detection further identified specific viral groups, namely, human adenovirus, polyomavirus and hepatitis A and C virus genomes present in the Umhlangane River. These viruses may impose a great health risk to individuals who may directly or indirectly utilize this water source. The findings presented in this study reiterate the potential danger of enteric viruses and bacterial-dependent phage persistence in water environments. Furthermore, land use activities and poor management in effluent disposal is a lesson not just for organizations dealing with the Umhlangane River but for all organizations managing any water resource, globally. Finally, a systematic integrated risk assessment and management plan to identify and minimize sources of faecal contamination is the most effective way of ensuring water safety and should be established in all future guidelines.
STUDY LIMITATIONS
This study identified a diverse array of virus particles in the Umhlangane River. However, some limitations were noted. Although seasonal effects were not observed, the short sampling period of 6 months did not cover a seasonal cycle. Therefore, changes that may have occurred at a later stage would not have been detected. Additionally, to fully assess the extent of viral infectivity and the multiplicity of infection tissue culture assays are required.
ACKNOWLEDGEMENTS
The authors would like to thank the National Research Foundation, South Africa for providing the Masters scholarship. The authors declare that they have no competing interests.