Virus isolations from sewage and from a stream receiving effluents of sewage treatment plants*

Bull. Org. mond. Sant} 1970, 42, 291-296Bull. Wid Hith Org.J Virus Isolations from Sewage and from a Stream Receiving Effluents of Sewage Treatment Plants* SAUL GRINSTEIN, JOSEPH L. MELNICK & CRAIG WALLIS In order to detect viruses in sewage or streams, it is first necessary to concentrate the virus present in the fluid sample. Available methods are not readily manageable for con- centrating virus from large volumes of fluid, and have not always yielded high recovery rates. In the study described in this paper, a method for concentration of viruses by ad- sorption on insoluble cross-linked maleic anhydride polyelectrolytes has been utilized to survey the viral flora of sewage and of a stream receiving sewage effluents, in a residential area of Houston, Texas. On a single day the virus flow at different points along the stream varied from 304 000 to 6 014 000 PFU/min. From 84 samples each of I US gal, 14 520 isolates were obtained, chiefly echovirus type 7 and polioviruses ofall 3 types, some ofthem with characteristics of virulent wild strains. With virus isolation rates as high as those achieved, it is now possible to monitor virus in natural waters more effectively. In order to detect viruses in sewage or streams, it is first necessary to concentrate the fluid sample. Available methods are not readily manageable for concentrating virus from large volumes of fluid, and have not always yielded high recovery rates (Gard, 1940; Gravelle & Chin, 1961; Kelly, 1953; Lund & Hedstrom, 1966; Metcalf & Stiles, 1968; Melnick, 1947; Melnick et al., 1954; Riordan, 1962). The method for concentration of viruses by adsorption on insoluble cross-linked maleic an- hydride polyelectrolytes (Wallis, Melnick & Fields 1) has been adapted to the detection of viruses in sewage and excreta (Wallis et al., 1969). Utilizing the polyelectrolyte technique, a survey was made of the viral flora of sewage and of a stream receiving sewage effluents, in a residential area of Houston, Texas, during the summer of 1968. MATERIALS AND METHODS Collection points and sampling methods The accompanying map shows the stream from which samples were taken (Brays Bayou), with the * From the World Health Organization International Reference Centre for Enteroviruses, Department of Virology and Epidemiology, Baylor College of Medicine, Houston. Texas 77025, USA. This study was supported in part by research grant Al 05382 from the National Institute of Allergy and Infectious Diseases, National Institutes of Health of the USA. 1 Paper in preparation. collection stations designated as point 1 (the collec- tion farthest upstream in the 10-mile (16-km) section of the stream under study) to point 11 (located immediately above the point at which the stream widens to become a river). At a distance of 4-5 miles (6.5-8 km) upstream from points 1 and 2, there were other sewage treatment plants which also empty into the stream, and residential sections con- taining a large number of septic tanks which drain into it. Collection stations 3, 5 and 8 were sewage treatment plants; here the samples were taken from the settling tank, at a step immediately preceding chlorination. Chlorination procedures (1 ppm) were similar for all 3 plants studied. A 1-US gal (3.8 litre) sample of sewage or stream water was collected (about 1 ft, i.e., 30 cm, below the fluid surface) at each of the points indicated. About 1 min was required to fill the collection bottle. The samples were returned to the laboratory and assayed for virus at once. Samples from the same collection points were obtained at weekly intervals (lUS gal at each point, taken at about 10 a.m.) during the months of April, May, July and August 1968. They were treated by use of the polyelectrolyte technique and tested for the presence of enteric viruses, as de- scribed below. In the month of June 1968, Houston had 20 days of heavy rainfall, and the treatment plants and the stream were not monitored for virus during this period. 2469 -291- 292 S. GRINSTEIN, J. L. MELNICK & C. WALLIS THE PART OF HOUSTON, TEXAS, MONITORED FOR VIRUS FLORA N~~~~~~~~~~~~.U BAYOU PL~T5 MACGRPLAR1 * TRAIL~'S 'AS o 1 2miles PLANoT 2 3km S RAW SEWAGE ALTREATED SEWAGE EEC~~~~~~~~~~~~~~~~~~N91810 For a single day's simultaneous sampling of all 11 points (28 July 1968), collections were made within a period of 1 hour. At the same time, with the assistance of city sanitary engineers, determinations were made of the fluid volumes of the sewage plants and the stream, and of the flow rates per minute, at each collection point. The flow rate at point 11 could not be determined because of the broadening of the stream. Polyelectrolyte method The development and description of the technique will be reported in detail elsewhere (Wallis, Melnick & Fields 1; Wallis et al., 1969). In brief, each US gallon (3.8 litres) of sewage or stream water was transferred to a 1-US gal aspirator bottle, to which was added 400 mg of an insoluble maleic anhydride polyelectrolyte (PE60).2 The pH was adjusted to 6.0, and the fluid was magnetically stirred for 1 hour at 25°C. The mixture was filtered through a coarse glass-fibre pad to trap the polyelectrolyte, which was then recovered from the pad with the aid of a spatula. The polyelectrolyte was then suspended in 3 ml-5 ml of 10% foetal calf serum in pH 9 borate buffer, which eluted virus from the polyelectrolyte IPaper in preparation. 2 Provided by the Monsanto Company, St Louis, Mis- souri, USA. (Wallis et al., 1969). The eluates were centrifuged at 2000 rev/min for 5 min and passed through 0.45-u serum-treated membranes to remove bacteria. The filtrates were than tested for virus as described below. Although previously we had shown that PE60 readily adsorbed a number of enteroviruses, reovirus, and adenovirus (Wallis, Melnick & Fields 1), it was necessary to ensure that, in sewage, PE60 remained capable of revealing a variety of enteric viruses rather than selectively adsorbing only certain of them. Experiments were therefore performed by addition of known amounts of echoviruses and coxsackie- viruses to sewage. In addition to the virus types that were isolated during this study, it was found that prototype echovirus types 1, 3, 9, 12, 13, 26 and 30, and coxsackievirus types B4 and B5, could be efficiently recovered from sewage by the method employed. Monkey kidney (MK) cells Kidneys from immature rhesus monkeys were trypsinized and grown in Melnick's medium A containing 2% calf serum. Isolation of viruses Each of the filtrates obtained by the polyelectrolyte method described above was inoculated on to drained MK cultures using 0.1 ml per bottle culture, VIRUS ISOLATIONS FROM SEWAGE and 10 cultures per sample. Cultures were sub- sequently overlaid with the agar medium. When plaques appeared (10-day observation period), virus was plucked from the plaques and passed to fresh cultures maintained under fluid medium. Progeny virus was subsequently identified by neutralization tests. Raw samples of the unconcentrated sewage were passed through 0.45-I, serum-treated membranes to remove bacteria, and 1.0 ml was tested as above, 0.1 ml per bottle culture. Overlay medium Overlay medium consisted of Earle's salt solution, amino acids as used in Eagle's medium, 0.4% NaHCO3, 1: 60 000 neutral red and 1.5% Bacto- agar (Difco). Neutralization tests Selected isolates were tested by use of combination antiserum pools (Lim & Benyesh-Melnick, 1960). Genietic markers Tests for characterizing polioviruses have been described in detail elsewhere (Benyesh-Melnick & Melnick, 1959). In general, the d marker test was conducted for poliovirus isolates by determining their efficiency of plaquing under acid or alkaline overlays. Attenuated strains are delayed in plaquing under acid overlays (d-), whereas virulent strains plaque efficiently (d+). The T marker (rct/40) test was used to determine the growth of poliovirus isolates at elevated temperatures (40.5°C). Attenuated viruses fail to grow at this temperature (T-) whereas virulent viruses grow effectively (T+). RESULTS Amount of virus present in stream and sewage in single-day survey As can be seen in Table 1, the concentration of viable virus recovered from various points in the stream varied only over an 8-fold range as far downstream as point 11. By referring to the accom- panying map, it can be seen that point 11 is at least 5 miles (8 km) downstream from the nearest point at which sewage effluent was fed into the stream. Although virus was present in higher concentrations in the pre-chlorination sewage taken at points 3, 5, and 8, the amount still present in the stream that received the effluent from each plant after chlorination demonstrates that the chlorination methods used are not viricidal. TABLE I SURVEY OF SEWAGE PLANTS AND THE STREAM RECEIVING SEWAGE EFFLUENTS Total virus recovered CalculatedCollection Flow rate from virus flow ratepoint a2 (US gal/mmn) 1-US gal virUsfownatsamples (PUmn(PFU/US gal) 1. Stream 822 150 123 000 2. Stream 3500 330 1155000 3. Plant 1 9 700 620 6 014 000 4. Stream 11 750 210 2467500 5. Plant 1' 1170 820 959400 6. Stream 13 000 200 2 600 000 7. Stream 13 500 150 2 025 000 8. Plant b 1 050 290 304 000 9. Stream 14 500 132 1 914 000 10. Stream 15250 130 1982500 11. River Not determined 110 - a See map and text. bi Collected from settling tank in plant, at the step preceding chlorination. As much as 2 467 000 PFU of virus per minute was flowing through this residential area (see map, point 4, and Table 1) on the day tested, and even 2 miles (3.2 km) downstream from point 8, where sewage effluent is received, the stream carried a virus flow of 1 982 500 PFU/min (at point 10). Virus isolates recovered during July and August In Table 2 is summarized the amount of virus isolated from this area in the 1-US gal samples taken each week during July and August 1968. From the 76 samples taken in July and August, 12 855 virus strains were isolated. The average number of isolates per collection station recovered each week throughout the period varied from 45/US gal to 286/US gal. Virus was present in virtually every sample taken at each collection point. It is evident from these results that the polyelectrolyte concentration procedures gave consistently high rates of detection. The 1-ml samples of unconcentrated sewage were also tested simultaneously with each concentrate, and yielded only 2 isolates. This is not surprising, for the concentration of virus would have to be 8 293 S. GRINSTEIN, J. L. MELNICK & C. WALLIS TABLE 2 RECOVERY OF VIRUSES PER WEEK DURING JULY AND AUGUST BY THE POLYELECTROLYTE METHOD Total no. of virus isolates per US gal a Sampling points July August Total(see map) Week I Week 2 Week3 Week 4 Week I Week2 b Week3 Week 4 1 141 51 75 150 54 12 0 10 493 2 381 81 99 330 282 9 78 42 1302 3 (Plant) NS 561 190 620 640 102 610 480 3 203 4 210 30 48 210 260 40 33 NS 831 5 (Plant) 250 605 110 820 520 160 480 340 3 285 6 NS 230 170 200 NS 40 NS NS 640 7 140 10 110 150 112 NS 30 NS 552 8 (Plant) 160 170 260 290 220 60 200 220 1 580 9 40 80 99 132 60 9 0 80 500 10 20 10 20 130 35 6 NS 9 230 11 27 30 33 110 39 NS NS NS 239 Total 1 369 1 858 1 214 3 142 2 222 438 1 431 1 181 12 855 Average 137 169 110 286 222 45 178 169 1169 a NS = not sampled. b Heavy rainfalls during this week. over 4000 PFU/US gal to be detectable without concentration. Despite the large number of positives, the highest amount of virus recovered from the polyelectrolyte concentrates proved to be 820 PFU/ US gal. When the isolations from the 8 samples taken in April and May 1968, and tested by the same pro- cedure (Wallis et al., 1969), are added to those shown in Table 2, a total of 14 520 isolates was recovered from sewage and stream water during the 4 months studied. Of these, 778 were selected for typing. Neutralization tests with Lim-Benyesh- Melnick combination antiserum pools (Lim & Benyesh-Melnick, 1960) readily identified 772 iso- lates. Of these, 33 were type 1 poliovirus, 84 were type 2, and 73 were type 3. Of 580 echoviruses, 576 were type 7; 2 were type 15; 1 was type 6* atnd 1 was type 33. Two adenoviruses were isolated and it was not possible to classify 6 of the viruses isolated. Genetic markers ofpoliovirus strains isolated The poliovirus strains were examined for d and T (rct/40) markers. The results of this test are shown in Table 3. It is evident from the large number of d-T- strains, especially for types 1 and 2, that many of the poliovirus isolates came from vaccinated persons. However, there were appreciable numbers of d+T+ isolates, which are the markers of TABLE 3 GENETIC MARKERS OF POLIOVIRUS ISOLATES Number of isolations Poliovirus Genetic in 1968 type markers April May July August1 d-T- 5 6 7 3 21 Type I d+T- 1 1 0 4 6 d+T+ 6 0 0 0 6 d-T- 25 2 10 3 40 Type 2 d+T- 6 5 18 1 30 d+T+ 3 4 7 0 14 d-T- 3 4 2 5 14 Type 3 d+T- 10 9 8 5 32 d+T+ 10 5 6 6 27 Totals 69 36 58 27 190 294 VIRUS ISOLATIONS FROM SEWAGE strains associated with virulence. Thus, of 33 type 1 strains tested, 6 were d+T+; of 84 type 2 strains tested, 14 were d+T+; and of 73 type 3 strains tested, 27 were d+T+. DISCUSSION With virus isolation rates as high as those ob- tained by the method used in this study (Wallis et al., 1969), it is now possible to evaluate virus naturally present in waters within a community. From 84 samples each of 1 US gal of sewage or stream water collected over a single summer, 14 520 isolates were made. The isolation rates are higher than in any studies previously reported (Gard, 1940; Gravelle & Chin, 1961; Kelly, 1953; Lund & Hed- strom, 1966; Melnick, 1947; Melnick et al., 1954; Metcalf & Stiles, 1968; Riordan, 1962; Wallis & Melnick, 1967a, 1967b). Concentration procedures are important. Samples of raw unconcentrated sewage tested simultaneously with the polyelectrolyte concentrates were negative except for 2 isolates. Of the 76 concentrates tested, 74 were positive. The concentrates each week yielded an average of 45 PFU-286 PFU of virus per US gallon. In order to be detectable without concentration, the virus content of sewage would have to be over 4000 PFU/US gal. Approximately 5% of the virus isolates were typed; the greatest number of isolates were echo- virus type 7 (74%), which from previous surveys appears to be endemic in Houston (Wallis & Melnick, 1967a, 1967b). The usefulness of the procedure is particularly evident from the findings on the polio- viruses that were recovered. Of the 190 strains of poliovirus identified, 33 were type 1; 84 were type 2; and 73 were type 3. Of the 33 type-I strains isolated, most had vaccine markers, but 6 had d+T+ genetic markers, indicative of virulent strains. It is noteworthy that the strains with the markers associated with virulence were isolated chiefly in April, while strains with vaccine virus markers were mostly isolated during May, July and August. Two cases of type 1 paralytic poliomyelitis-one of them fatal-occurred in unvaccinated infants in Houston during June 1968, and a type 3 paralytic case occurred in January 1969. These were the first reported cases in the city since 1962 (Melnick et al., 1969). Of the 84 type-2 strains isolated, most again had vaccine markers, but 14 manifested d+T+ markers. In the tests of type 3 polioviruses, 27 of the 73 strains were d+T+. Since the type 3 vaccine strain is known to revert more readily toward d+T+ markers, it is difficult to evaluate whether or not the high propor- tion of type 3 isolates having these markers repre- sented strains derived from vaccine virus, or wild strains of this serotype (Benyesh-Melnick et al., 1967; Bull. Wld Hlth Org., 1969). The importance of inactivating viruses in sewage effluents is evident. Communities may release treated sewage into streams which serve as a source of drinking water for communities downstream. Sewage chlorination as currently practised does not inactivate enteroviruses, as was again made clear in this current study, since pre-chlorination influent samples and chlorinated effluent samples did not differ significantly in virus concentration. Infectious hepatitis is another enteric virus which, when present in sewage, has been found to resist chlorination (Melnick, 1957). As indicated by the findings of this study, a simple and efficient method is now available for detection of viruses in large volumes of sewage and streams by adsorbing viruses on insoluble cross-linked maleic anhydride polyelectrolytes, and then eluting them from these polymers into small volumes for assay. These methods are being applied in a search for virus in other natural waters. ACKNOWLEDGEMENTS The authors wish to thank Reuben Wende for assistance in setting up the project in the field, Jon Stem and Zane Segal for technical assistance, Marjorie Burkhardt for typing the viruses, and Verle Rennick for assistance with the preparation of the manuscript. 295 296 S. GRINSTEIN, J. L. MELNICK & C. WALLIS RtiSUMP- ISOLEMENTS DE VIRUS DANS DES EAUX D'EGOUT ET DANS UN COURS D'EAU 00 SE DEVERSENT LES EFFLUENTS D'USINES DE TRAITEMENT DES EAUX USEES Une nouvelle methode fondee sur la concentration des virus par adsorption sur un polyelectrolite insoluble (en l'occurrence un copolymere a liaison croisee de l'anhy- dride maleique de l'isobutylene) a et appliquee a l'iso- lement des virus dans des eaux d'egout et dans un cours d'eau oiu se deversent les effluents d'usines d'epu- ration. Durant l'Fte 1968, ai Houston (Texas), on a procede a des prelevements en 11 points d'un cours d'eau d'environ 16 km de long recevant les effluents de stations d'epu- ration. A partir de 84 echantillons de 1 US gallon (3,78 litres), on a effectue 14 520 isolements de virus. En un seul jour, les concentrations de virus debitees a differents points du cours d'eau variaient de 304 000 'a 6 014 000 unites formatrices de plage (PFU) par minute. Les taux d'isole- ment etaient les plus eleves jamais signales. Les echantil- Ions d'eau d'egout brute examines simultanement sans concentration pr6alable ont ete trouves negatifs a deux exceptions pres. Par contre, sur les 76 echantillons soumis a concentration, 74 etaient positifs. Le nombre moyen d'isolats recueillis chaque semaine a chaque point de collecte a varie entre 45 et 286 par gallon pendant toute la periode consideree. Pour pouvoir etre decelable sans concentration prealable, il faudrait que la teneur en virus des eaux d'egout depasse 4000 PFU par gallon. On a pu etablir le type d'environ 5% des virus isoles: il s'agissait essentiellement de poliovirus dqs types 1, 2 et 3 et d6chovirus du type 7. Sur les 190 souches de polio- virus isolees, 33 appartenaient au type 1, 84 au type 2 et 73 au type 3. La majorit6 des souches du type 1 posse- daient les caracteres genetiques des souches vaccinales, mais 6 d'entre elles possedaient les marqueurs d + et crt/40 + caracteristiques des souches virulentes. II vaut la peine de noter que ces dernieres ont ete isolees en avril, tandis que les premieres l'ont ete surtout en mai, juillet et aouit. A Houston, en juin 1968, deux nourrissons non vaccines ont ete atteints de poliomyelite paralytique du type 1 et l'un d'eux est mort. II y a eu aussi un cas de poliomyelite paralytique du type 3 en janvier 1969. C'6taient les premiers cas observes a Houston depuis 1962. La plupart des souches du type 2 possedaient des marqueurs caracteristiques des souches vaccinales, mais 14 avaient des marqueurs d + crt/40 +. Parmi les 73 isolats du type 3, 27 etaient marques d + crt/40 +; comme on sait que la souche vaccinale du type 3 est plus susceptible de reversion vers ces caracteres, il est difficile de determiner si ces isolats d + crt/40 + sont des souches deriv6es du virus vaccinal ou des souches sauvages de ce serotype. Les isolements de virus dans les effluents d'usines d'epuration viennent a l'appui d'observations anterieures indiquant que la chloration des eaux usees telle qu'elle se pratique actuellement n'inactive pas les ent6roNirus. Les concentrations de virus dans des echantillons d'eaux usees preleves avant la chloration et dans des effluents chlores ne presentaient pas de difflrence significative. REFERENCES Benyesh-Melnick, M. & Melnick, J. L. (1959) In: Live poliovirus vaccines. First International Conference on Live Poliovirus Vaccines... 1959, Washington, D.C., Pan American Sanitary Bureau, pp. 179-202 Benyesh-Melnick, M., Melnick, J. L. Rawls, W. E., Wimberly, I., Barrera Oro, J., Ben-Porath, E. & Rennick, V. (1967) Amer. J. Epidem., 86, 112-136 Bull. Wld Hlth Org., 1969, 40, 295-300 Gard, S. (1940) J. exp. Med., 71, 779-785 Gravelle, C. R. & Chin, T. D. Y. (1961) J. infect. Dis., 109, 205-209 Kelly, S. M. (1953) Amer. J. publ. Hlth, 43, 1532-1538 Lim, K. A. & Benyesh-Melnick, M. (1960) J. Immunol., 84, 309-317 Lund, E. & Hedstrom, C.-E (1966) Amer. J. Epidem., 84, 287-291 Melnick, J. L. (1947) Amer. J. Hyg., 45, 240-253 Melnick, J. L. (1957) A water-borne urban epidemic of hepatitis. In: Hepatitis frontiers (Henry Ford Hospital International Symposium), Boston, Little, Brown & Co., pp. 211-225 Melnick, J. L., Burkhardt, M., Taber, L. H. & Erckman, P. N. (1969) J. Amer. med. Ass., 209, 1181-1185., Melnick, J. L., Emmons, J., Opton, E. M. & Coffey, J. H. (1954) Amer. J. Hyg., 49, 185-195 Metcalf, T. G. & Stiles, W. C. (1968) Amer. J. Epidem.; 88, 379-391 Riordan, J. T. (1962) Yale J. Biol. Med., 34, 512-521 Wallis, C., Grinstein, S., Melnick, J. L. & Fields, J. E. (1969) Appl. Microbiol. 18, 1007-1014 Wallis, C. & Melnick, J. L. (1967a) Amer. J. Epidem., 85, 459-468 Wallis, C. & Melnick, J. L. (1967b) Bull. Wld Hlth Org., 36, 219-225