Isolation of viruses from sewage, with special regard to poliovirus

Bulletin of the World Health OrganIzation, 56 (6): 937-943 (1978) Isolation of viruses from sewage, with special regard to poliovirus An evaluation of the sensitivity of three tissue culture methods MARGARETA B6TTIGER1 This report concerns experiments to isolate different viruses from sewage. Using a special cell-line from Utrecht, derivedfrom human amniotic cells, it was possible to isolate poliovirus selectively when antisera against six types of coxsackievirusB were added to the tissue culture. The method was tested in connexion with the epidemiological investigation ofa case ofpoliomyelitis in Sweden in 1977. It rapidly demonstrated that the virus implicated was present in all neighbouring sewage plants, indicating a wide distribution of the virus in the area. This study was a continuation of earlier trials that dealt with the isolation of viruses from sewage with special regard to poliovirus (1). Sewage was found to contain many different viruses and it was considered unsatisfactory simply to determine the presence of one or more specific virus types by inoculation of sewage into a conventional, relatively broad-spectrum, cell-culture system. Even when isolation was repeated in the presence of hyperim- mune sera to the first identified strains this procedure was too unreliable, time-consuming, and impractical for general use in the evaluation of viral contamina- tion of sewage. Selective methods of isolation were therefore examined. At first, three different cell-culture systems were tested; but as the main object was to isolate poliovirus, a more complex system using different cell cultures in combination with pools of hyperimmune sera was also tested and, later, a fourth tissue was introduced. Finally, by adding known amounts of poliovirus to the sewers, an attempt was made to evaluate the sen- sitivity of the methods. MATERIAL AND METHODS Collection ofsewage Samples of sewage were collected at the inlet of the sewage plant by either using gauze swabs or taking crude material from the screen as described earlier (1). 1 National Epidemiologist, Epidemiological Department, National Bacteriological Laboratory, 105 21 Stockholm, Sweden. Methods of virus isolation The treatment of the sewage samples before inoculation was the same as in the earlier study (1) mainly following the recommendations of Riordan (2). In brief, after making the raw sewage alkaline (to pH 9) and compressing it, the bacteria were separated from the sewage by chloroform treatment and centrifugation. The pretreated extract was inoculated into three or four different types of tissue culture (5 ml into each culture) in 200-ml bottles. The cell cultures used were primary cyno- molgus monkey kidney cells (Cyn), green monkey kidney cells (GMK), a human amnion cell line from Utrecht (U), and diploid human embryonic lung cells (HL). The isolation studies were carried out in two steps. The first study was performed with 19 extracts collected in the autumn over a period of about two months. The different steps of the isolations are illustrated in Tables 1-3. Thus, initially three different tissues were inoculated (Cyn, GMK, and U). After one or two weeks the cell and medium mixtures were passed to four different tube cultures-Cyn, GMK, U, and HL. Typing of virus was performed in cells of the same kind as the ones used for the second passage. Pools of 5-6 hyperimmune sera against the following viruses were used: echovirus types 1-30, coxsackievirus A type 9, coxsackievirus B types 1-6, poliovirus types 1-3, and reovirus types 1-3. The technical procedures used were the same as those described earlier (1). As a rule, easily distinguishable virus isolates were typed only once or twice. Reisola- 3762 -937- M. BOlTIGER Table 1. Results of isolation attempts involving U cells a U U Cyn GMK U GMK Cyn HL U Cyn U GMK U GMK Cyn GMK Cyn HL U U U U Poliovirus 2 2 2 2 2 2 0 0 0 0 Echovirus 0 0 0 0 0 5 0 0 0 0 Coxsackievirus B 18 16 10 16 18 16 7 5 8 4 Reovirus 0 0 0 0 0 0 0 0 0 0 Novirus 0 0 0 0 0 5 11 9 6 15 Total no. of samples tested 19 18 10 18 18 16 18 14 14 19 a Types of viruses disclosed by the first attempt of typing by use of neutralization tests with groups of pooled hyperimmune sera. Altogether 19 samples were included in the study but all were not tested on all types of cell cultures. The 2 samples containing poliovirus were included in all tests.U - human amnion cell line; HL - human diploid embryonic lung strain; Cyn - monkey kidney primary cell culture; GMK - green monkey kidney cell line. Table 2. Isolation attempts involving the two monkey tissues only a Cyn Cyn GMK GMK Cyn GMK GMK Cyn Cyn GMK GMK GMK Cyn Cyn Poliovirus 0 0 0 0 0 0 Echovirus 13 13 10 15 (7) 13 Coxsackievirus 4 4 2 3 (2) 4 Reovirus 2 3 2 2 (1) 3 No virus 1 1 1 1 (1) 0 Total no. of samplestested 18 18 19 19 (10) 17 a Symbols as in Table 1. tion in the presence of type-specific hyperimmune serum was not carried out. Typing within pools was perforned only in a small number of the isola- tion trials. In a second study, freshly collected extracts were submitted to the isolation procedure presented in Table 4. The extracts were inoculated directly into the four different types of cell culture. Each type of culture was passed to cells of the same type. The U cells were also inoculated with an extract which had been allowed to bind for 1 hour at 370C with a pool of hyperimmune sera to the six types of coxsackievirus B. This latter culture was further passed, on the one hand, in the same manner as before with hyperimmune serum and, on the other hand, to cultures of human diploid cells. Typing of isolates was performed as earlier and in cell cultures of the same kind as used for isolation. Thirdly, successively reduced amounts of atten- uated poliovirus (the Chat strain) were added to the sewage system about 4.5 km above the inlet and reisolation of the virus was attempted at the inlet. 938 ISOLATION OF VIRUSES FROM SEWAGE Table 3. Results of isolation attempts involving human lung fibroblasts in the second passage a GMK Cyn HL HL HL Cyn GMK HL C n GMK Poliovirus 0 0 0 0 0 0 Echovirus 14 9 14 11 11 7 Coxsacklevirus 1 1 3 1 1 0 Reovirus 0 1 1 0 1 1 Novirus 3 1 1 3 3 2 Total no. of samples tested 18 12 18 15 16 11 a Symbols as in Table 1. Table 4. Results of continued isolation trials of 12 sewage samples on different cell cultures and with a pool of antisera to coxsackievirus B added to the U cell cultures a U pool of immune sera to Sample GMK U coxsackievirus HL no. / GMK U HL U HL HL 1 coxsackie B coxsackio B (+) + (+) echo 2 + coxsackis B (+) - (+) + 3 coxsackle B coxsackie B (+) - + echo 4 coxsackie B coxsackie B (+) - + echo 5 echo . coxsackie B (+) - (+) echo 6 coxsackie B coxsackie B + + + + 7 coxsackie B coxsackie B (+) - (+) (+) 8 coxsackle B coxseckie B (+) + + echo echo 9 coxeackie B coxeackle B n.d. - n.d. (+) 10 coxsackie B coxeackle B n.d. (+) n.d. echo 11 polio coxeackie B (+) polio polio echo 12 echo coxsckie B n.d. polio polio echo a + * rapid growth, not typed or typing failed; (+) - slow replicating virus, not poliovirus; - - no virus growth; n.d. - not done. 939, M. BOTTIGER Using a virus suspension containing 10 TCID60 per ml, the following amounts of suspension were added: 6000, 1000, 200, 40, and 8 ml. The probable time interval between the addition of virus to the stream and its entry into the sewage plant was evaluated by isotope studies; the dispersion of isotope activity was also evaluated. All tests were carried out under normal circumstances and at the same time of day, i.e., in the morning. Sewage samples were collected at the inlet before virus was added, at the time when virus was likely to be entering the plant, and 24 hours later. RESULTS Isolations and passages involving U cells are illustrated in Table 1. It can be clearly seen that in all isolations involving U cells, with one exception, only coxsackievirus B viruses and polioviruses were isolated. The exception occurred when isolates were passed directly from U cells to human diploid lung cells (HL). When isolates were passed from U cells to monkey cells (GMK and Cyn) the coxsackie B viruses quickly became dominant, but on passage to HL cells a number of echoviruses were detected, which appar- ently had survived the first passage on U cells. The coxsackieviruses isolated were mainly of types 3 and 5. Systematic testing of the ability of enteroviruses to multiply on the U cells demonstrated that of the echoviruses only type 11 would multiply, and that all the coxsackie B viruses generally grew out, especially types 3 and 5. Poliovirus was isolated from two samples and was easily detected in sub- sequent passages on all the other tissues. Isolations involving primary cynomolgus monkey cells and a green monkey kidney cell line The two poliovirus isolates detected when U cells were used as the primary isolation tissue were not found when the two monkey kidney tissues were used for this purpose. Various echoviruses were dominant among the viruses first isolated and were readily detected in about 75%O of the samples. Although all the samples contained coxsackieviruses, as was shown by the U cell isolation tests, they were apparently inhibited or overgrown by the echo- viruses. Reoviruses were typed in a small number of the isolates (2 or 3 out of 19 samples). When GMK cells were used for primary isolation the subtyping of echoviruses within groups of pooled sera revealed that in 8 instances out of 9 the first strain detected was echovirus type 7. On the ninth occasion echovirus type 1 was detected first. When isolation was performed on cynomolgus cells, echovirus type 1 replicated more readily from the same set of samples. Five type 1 and three type 7 echoviruses were found. Echovirus type 18 was also found in one isolate when primary cynomolgus cells were used as the first step in the isolation procedure. When samples inoculated into monkey cells (of both kinds) were passed to human lung cells, viruses belonging to the echo group were again selected (Table 3). Trials involving the use of hyperimmune sera to coxsackie B viruses and the testing of human lung cells for the primary inoculation The results of further selective isolation studies on 12 new samples are given in Table 4. Some of the trends found in the first study were repeated. Thus, coxsackieviruses were isolated from all passages on U cells. Cultures in GMK cells revealed a mixture of echoviruses and coxsackieviruses. Poliovirus was found in two samples (11 and 12). In one sample (12), poliovirus was detected only by passage in U cells in combination with anticoxsackie B serum. In the other, it was also found by passage in GMK cells. Studies of reisolation of attenuated poliovirus vaccine added to the sewers In the first trial, 6 litres of virus suspension, containing 106 TCID50 of virus per ml, were poured into a main sewage tunnel at a distance of about 4.5 km above the inlet of the sewage plant. As the rate of flow was estimated to be about 1 m/s, samples were collected at different times from about 1-11/2 h after the addition. One month later the trial was repeated using only 1 litre of virus suspension of the same batch. In these two first trials only GMK tissue was used to reisolate the virus from the sewage (Table 5, top). It was possible to detect the virus on both occasions. Further tests were then carried out to evaluate the sensitivity of the method, i.e., to find the minimum amount of virus that it was possible to detect. The flow rate of the sewage was also evaluated with radioactive isotopes and it was found that the peak of radioactivity at the inlet was reached about 60 min after addition of the isotopes 4.5 km from the inlet. Activity was first detected about 55 min after and had disappeared 70 min after adding the isotopes; during this period the mean rate of flow of sewage was 2.4 ms per second. In the light of this information, the reisolation experiments were 940 ISOLATION OF VIRUSES FROM SEWAGE Table 5. Results of trials with attenuated poliovirus type 1 added to the sewers 4,5 km above the inlet of the sewage plant Tissue used for Volume of vaccine Virus isolation from swabs:Test ioaon suspension seededisolation (106 TCIDao per ml) before seeding after 1 % h after 1 W-24 h I GMK 6000 ml CPAa Poliovirus type 1 Poliovirus type 1 2 GMK 1000 ml CpAa Poliovirus type I coxsackievirus, echovirus reovirus 3 GMK 200 ml coxsackievirus Poliovirus type 1 coxsackievirus u coxsackievirus (mixture ?) coxsackievirus U + ISCb coxsackievirus Poliovirus type 1 negative HL Poliovirus type 1 echovirus GMK 40 m coxsackievirus coxsackievirus coxsackievirus u coxsackievirus coxsackievirus coxsackievirus + poliovirus type 1 u + ISC b negative Poliovirus type I negative HL CPA a echovirus 5 GMK 8 ml echovirus coxsackievirus echovirus u CPAa coxsackievirus coxsackievirus, echovirus, poliovirus type 1 U + ISCb CPAa negative CpAa H L CPA'a echovirus echovirus a CPA = cytopathogenic agent, not poliovirus. b ISC = immune sera against coxsackieviruses. repeated using different tissue cultures and the addition of immune sera to coxsackieviruses as described earlier. The results of the three later tests are described in Table 5 (tests 3-5). After the addition of 200 ml of virus suspension to the sewage flow, virus could be readily reisolated in GMK, U cells with anticoxsackie B serum, and HL cells, from a swab that had been hanging at the inlet. Virus was not detected in 5 ml of ordinary sewage water. When 40 ml of virus suspension was added to the main sewage pipe, virus was detected at the inlet after 1½/2 hours, but only in U cell culture. Poliovirus was also found in the swab immersed in the sewage 11/2-24 hours after the introduction of the virus. The last test, using only 8 ml of virus suspension, indicated that even this small amount of virus could be detected. Again the U cell cultures appeared to be the strain of choice for selective isolation of polio- virus. The first real opportunity to test the practical application of these investigations in a real situation occurred in February 1977, when the first indigenous case of poliomyelitis in Sweden since 1962 was diagnosed (3). The patient belonged to a special community, which does not approve of vaccination. It was therefore of epidemiological interest to investigate quickly the spread of the virus in the community. The incubation of sewage samples with coxsackie- virus B antiserum followed by inoculation on cultures of U cells revealed the presence of poliovirus in all the neighbouring sewage stations and confirmed the wide distribution of the virus. DISCUSSION Although selective cultures for the isolation of bacteria have been widely used for many years, no such routine procedures have been devised for virus isolation. This is an important problem when a mixture of viruses is obtained, for example, from sewage waters and it is necessary to study the epidemiology and spread of specific viral enteric diseases of man or animals. With virus isolation from sewage there is an additional problem. Before the samples can be inoculated into cell cultures, other contaminants such as bacteria and toxic substances must be eliminated, and care must be taken not to destroy the viruses. The procedure is complicated by the fact that different cell cultures react differently to these toxic agents. This preliminary study was concerned primarily with developing a quick method of isolating polio- virus from sewage. Sweden, Finland, and Holland are the only countries practising immunization against poliomyelitis using only inactivated virus 941 M. BOTTIGER vaccine and thus the situation in Sweden is of particular interest since the sewage is not contamin- ated by live vaccine viruses. After using killed vaccine for 20 years it is of the utmost interest to know how much and what kind of polioviruses circulate in the country. In countries practising oral vaccination we now know that the majority of the strains isolated are vaccine derived (4). Both green monkey kidney (GMK) and primary cynomolgous monkey kidney (Cyn) cells are commonly used for isolation of enteroviruses. The present studies confirmed that, in the absence of other strains, polioviruses will grow very well in these cells, but it was noted that when coxsackie- viruses and echoviruses were present they interfered significantly with the isolation of poliovirus. In an earlier study by the author (1), performed with Cyn cells, echovirus type 1 was readily isolated. In the present investigations there were some indications of a difference between the GMK and Cyn cultures. In the GMK cells, echovirus type 7 seemed to replicate more readily than echovirus type 1; other- wise they appeared to have a similar sensitivity to poliovirus and coxsackievirus infection. The U cells were unquestionably the best for isolation of coxsackieviruses and polioviruses. No sewage sample collected from swabs or crude material from the screen has so far been negative for coxsackievirus when tested in these cells. The readiness with which this virus grows interferes with polioviruses if they are present in only small amounts. This is demonstrated quite clearly by the results of isolations from samples 11 and 12 (Table 4). On pure U cell cultures coxsackievirus was first detected. On U cell cultures to which hyperimmune sera against coxsackievirus B strains had been added, however, polioviruses readily grew out. The human diploid cell cultures revealed mainly echoviruses (Tables 1, 3, and 4). When there was a competition between polioviruses and echoviruses the latter seemed to dominate. The combination of U cells for primary isolation and HL cells for passage also appeared appropriate for the isolation of poliovirus. The first tissue favoured coxsackieviruses and polioviruses, the second was also sensitive to polioviruses but had relatively little susceptibility for the coxsackie- viruses. Although this combination of two cell systems, using first one and then the other, appeared possible for selection of poliovirus, the use of U cells in combination with the pool of immune sera to coxsackieviruses seemed to be the most sensitive and most rapid method for the selective isolation of poliovirus from a mixture containing numerous other viruses. The simple sampling method using only a simple swab provided dramatic evidence of the sensitivity of the isolation methods. When 40 ml of virus suspension was added to the sewer, i.e., 4 x 107 TCID50 of virus, and if we assume that about 2 x 106 litres of sewage water passed by during 15 min, 1 litre would have contained on average 20 TCID6o of virus. In this rough calculation no allowance has been made for adsorption of virus particles to the tunnel or to other material or for inactivation. The selective method for quick isolation of polio- virus alone from sewage showed itself to be most valuable when a case of poliomyelitis was diagnosed in 1977 (3). It was readily demonstrated that the virus was widespread in the community and that general vaccination of unprotected persons was indicated. RESUME ISOLEMENT DES VIRUS PRiSENTS DANS LES EAUX USiES, NOTAMMENT LES POLIOVIRUS: EVALUATION DE LA SENSIBILITi DE TROIS METHODES UTILISANT DIFFERENTES CULTURES CELLULAIRES Par l'immersion de tampons dans un egout collecteur ou ils s'imbibent d'eaux us6es, on s'est procure les echan- tillons n6cessaires a la detection d'agents cytopathogenes. Plusieurs cultures cellulaires ont et6 utilis6es a cette fin: une primoculture de cellules renales de singe cynomolgus (Cyn), une lignee cellulaire obtenue a partir du rein de singe vert (GMK), une lignee de cellules amniotiques humaines (U) et des fibroblastes diplolides humains pro- venant d'un poumon embryonaire (HL). On a constate a la faveur d'une serie d'exp6riences que, pour un meme echantillon d'eaux usees, differents types de virus etaient deceles selon la culture cellulaire utilisee. Ainsi, on a isole principalement des virus Coxsackie et des polio- virus a partir des cultures de la lignee cellulaire amnio- tique (U) inoculees. La presence d'echovirus a ete r6v6l6e au moyen de cultures de fibroblastes pulmonaires (HL); 942 ISOLATION OF VIRUSES FROM SEWAGE 943 enfin, les cultures de cellules renales de singe (GMK et Cyn) ont mis en evidence la presence de melanges de tous ces virus. Apres avoir combine les cultures de cellules amnio- tiques (U) avec du serum anti-Coxsackie, on a pu isoler des poliovirus rapidement et d'une maniere selective. Cette methode s'est montr6e la plus efficace dans une situation d'urgence - c'est-a-dire lorsque les risques d'epidemie lids a la propagation de poliovirus sp6cifiques doivent etre 6valu6s dans le plus bref dMlai - et elle s'est aussi revelee extremement sensible. Ceci a ete verifie en introduisant dans un egout collecteur une suspension de poliovirus attenues titrant 106 TCID5o par ml a une distance de 4,5 km en amont de la station d'epuration. Le d6bit moyen total 6tait de 2,4 m3 par seconde a l'arrivee, oui le virus a ete effectivement retrouve. L'etude montre qu'il est possible de surveiller la dis- semination par les eaux usees d'agents viraux pathogenes, en depit d'eventuelles complications - dues notamment a la presence dans ces eaux de tres nombreux types de contaminants et aux interferences qu'elle entraine. C'est pourquoi il importe de mettre au point des methodes d'isolement de ces agents aussi selectives et sensibles que celles qui sont couramment employees dans des conditions semblables pour l'identification des bacteries. REFERENCES 1. BWrrIGER, M. Experiences from investigations of virus isolations from sewage over a two year period with special regard to polioviruses. Archiv far die gesamte Virusforschung, 41: 80-85 (1973). 2. RIORDAN, J. T. Isolation of enteroviruses from sewage and after vaccine administration. Yale journal of biology and medicine, 34: 512 (1962). 3. BOTTIGER, M. ET AL. Bulletin of the World Health Organization (in press). 4. COSSART, Y. E. Evolution of poliovirus since introduc- tion of attenuated vaccine. British medical journal, 1: 1621-1623 (1977).