Reviews

UDC 614.777:579.841.11]-078

G. P. Kalina, N. B. Komzolova

Pseudomonas aeruginosa and the Hydrosphere: Surface Waters

F. F. Erisman Moscow Research Institute of Hygiene

In 1902, A. Houston [34] proposed including Pseudomonas aeruginosa (the silver pseudomonad, hereafter PA) among the indicators of surface-water contamination by sewage. This recommendation, however, was not adopted [10], and studies of water were sporadic. Several decades passed before the issue gained importance in light of compelling evidence of the high pathogenicity of PA [4]. Illnesses among people who had come into contact with PA-contaminated water during recreational activities [26], in swimming pools [29], and while deep diving [8] provided decisive grounds for recognizing the role of water in the epidemiology of pseudomonas infections [46].

It has been claimed [29], overlooking A. Houston, that the first reliable report of the isolation of PA from water was published in 1956 [36]. E. Taylor [60], however, noted that outbreaks of waterborne diarrhea had been recorded in the United States and England as early as the turn of the twentieth century, with PA isolated from both the patients and the water. The development of specialized selective media in the early 1960s marked the beginning of systematic and wide-ranging investigations of PA in different types of water.

The information available to us cannot be fully covered within the scope of this review. The table presents only data that, in our view, are most significant or characteristic with respect to the biotopes, the research methods used, or the results obtained. Numerous…

Presence of Pseudomonas aeruginosa in surface waters

Note. MPN/100, most probable number per 100 mL; VMC/100, number of viable microbial cells per 100 mL; %/\(X\), percentage of positive samples when a volume \(X\) of water was inoculated; F, membrane filter. Media: A, liquid (A1, standard glucose peptone; A2, synthetic medium with pyocyanin [52]; A3, selenite medium diluted twofold with the water under investigation [45]; A4, malachite green solution, concentration unspecified [23]; A5, asparagine–ethanol medium [21]; A6, Hedberg’s acetamide medium; A7, synthetic medium with sodium ammonium citrate [37]; A8, citrate–penicillin medium [10]); B, solid (B1, standard nutrient agar; B2, MacConkey agar; B3, modified Endo agar [39]; B4, King B medium; B5, nutrient agar with TTC [55]; B6, mPA [42]; B7, nutrient agar with propanid [1]; B8, mPA-B [22]; B9, mPA-C [14]; B10, acetamide agar; B11, asparagine–ethanol agar [21]; B12, synthetic medium with ferric ammonium citrate [37]). F and S denote fast and slow flow, respectively.

…reports on individual aspects of the problem.

Various authors assessed the degree of water contamination with СП by density (most probable number, MPN), the number of viable microbial cells (VMC in \(N\) volume of water), and frequency of detection (the percentage of samples in which СП was detected). Some investigators made only a qualitative determination on a presence-or-absence basis [18, 35].

A critical analysis of the material available to us raised two questions to which more than one answer is possible: 1) What is the degree of water contamination with this pathogen, and at what level does it become sufficient to constitute a significant epidemiological factor? 2) Is СП an autochthonous microbial inhabitant of water [50], or is it an allochthonous inhabitant [53] (if the latter is true, what is its ecological niche, and how often is water contaminated with this pathogen)?

The data presented in the table show considerable variability. This may be attributable to objective and subjective causes. The former include factors unrelated to human activity (climate, temperature, season, the physical state and chemical composition of the water, and the biocenosis of the biotope) or arising from human activity and collectively termed anthropogenic factors. The latter include the research methods used, the choice of culture media, the method of enumeration, and sample volume. These require critical analysis.

The correct choice of research method is essential for success and for obtaining the most reliable results possible. Studies of СП have used noninhibitory (routine) or selectively inhibitory media (with respect to Gram-positive microflora), inhibitory and selective media with a limited salt composition and various carbon and nitrogen sources, as well as combinations of liquid enrichment media and solid isolation media. Membrane filters have been used as an alternative to enrichment media. Parallel testing of two types of media or combinations thereof has made it possible not only to determine their effectiveness but also to obtain the most reliable results. In the absence of such a comparative assessment, differing biotope characteristics may lead to erroneous conclusions. The following examples demonstrate the importance of the method chosen. In 1966, a study of 1,065 water samples from small rivers in Hungary used filtration of 50 ml of water followed by incubation of the filter on modified Endo medium. СП was detected in 1.8% of the samples [39]. Three years later, 37,291 samples of Danube water were examined by a similar method, and only 0.5% yielded a positive result. In parallel, 1,372 samples of the same volume were inoculated into selenite enrichment medium diluted twofold with the water under study; growth of СП was obtained in 86.2% [45].

In the early 1960s, G. Bonde [10] investigated river water within Copenhagen for СП. Direct inoculation of 29 samples of heavily polluted water onto King B medium yielded negative results. Use of a synthetic medium with a limited salt composition, sodium citrate, and 2 IU of penicillin per ml permitted detection of СП in 9 of 23 samples. When liquid and solid media were combined according to the principle of Bonde medium, modified by the addition of further ingredients [37], СП was not detected in any of 32 river-water samples.

We tested Bonde medium at two sites. Initial testing of water from a small river with favorable sanitary characteristics yielded positive results in five 20-ml portions (pseudomonad MPN greater than 5/100). In three series of experiments, Bonde medium, with penicillin replaced by the less inhibitory crystal violet, and Drake’s asparagine–ethanol medium [21] were tested in parallel. Pseudomonad MPN was determined in two series. The value was 5.8/100 for Drake’s medium and 13.1/100 for Bonde medium. The results were practically equivalent, but Bonde medium is simpler and contains no scarce ingredients. These findings were fully confirmed in a study of 27 water samples from a large northern river (six samples were also inoculated in parallel into medium containing TTX) [55]. СП was detected in 63% of samples using Bonde medium. СП was isolated at a similar rate (66.7%) on inoculation into TTX medium, but the mean СП index was higher with Bonde medium than with TTX medium (14.8 and 3.7 per 1 l, respectively).

Comparison of these data with results obtained using the conventional sanitary-bacteriological method [2] to examine water from a river with similar characteristics in the same region (10.5% positive samples) showed the superiority of Bonde medium.

In 1972, the solid selective medium mPA [42] was developed in the United States on the basis of XLD medium, incorporating four inhibitors: sulfapyridine, kanamycin, actidione, and nalidixic acid. The medium’s effectiveness was low, although somewhat higher than that of Drake’s medium (see table). The basic composition was later modified slightly, and the sulfapyridine concentration was reduced (mPA-B medium). This improved the medium’s effectiveness, but optimal detection of СП required at least four days [22]. In Canada in 1978, mPA medium was rejected; removing sulfapyridine and actidione from mPA-B medium (mPA-C medium) allowed results to be read after just one day, with the same effectiveness as mPA-B medium—1,735 and 1,839 VMC/100, respectively [14]. Two of the four inhibitors proved unnecessary.

The importance of inhibitors in media used to isolate СП cannot, of course, be dismissed entirely. The addition of propanide to meat-peptone agar yielded fairly high results in studies of water from a large river with varying degrees of pollution [1], although a comparison of the propanide medium with inhibitor-free arginine and acetamide media for the examination of wastewater favored the latter [3, 5].

The effectiveness of СП isolation also depends on the volume of the sample examined, especially when water contamination is assessed by the most commonly used method, the percentage of positive samples. The negative results obtained by L. Ringen and C. Drake [52] may be explained by the fact that they analyzed very small volumes (1–2 ml) of clean mountain-stream water. In a study of water from a lake in the Federal Republic of Germany with a low ЖМК value per liter, СП were detected in almost all samples when the sample volume was 1,000 ml [56]. The most commonly used sample volume of 50–100 ml is sufficient for analyzing moderately polluted water, but cleaner waters should be examined in larger volumes (up to 1,000 ml). Using a method for the simultaneous detection of salmonellae and СП [35] in 5 liters of liquid, СП were found in 17 of 20 clean-water samples and salmonellae in 2.

A discrepancy between the detection of СП and the sanitary quality of water as determined by conventional assessments may be revealed when effective testing methods using selective media are employed. Hungarian researchers using an effective method noted that approximately half of the drinking-water samples that tested positive for СП would have been considered satisfactory on the basis of their coliform bacterial content [45]. In four ponds used intensively for bathing, P. aeruginosa was consistently detected despite low levels of fecal coliforms [47]. In the latter example, the extreme conditions—a large number of bathers, the absence of chlorination, the presence among bathers of people with ear, nose, and throat diseases, and possibly the antagonistic potential of СП—explain the predominance of СП over fecal coliforms. Evidence that antibiotic properties, in addition to the possibly greater survival of СП, are also important comes from reports of inhibition of metabolic processes in E. coli when grown together with СП on Endo medium [16], and of inhibition of Campylobacter growth by СП metabolites [9].

There are also contrary reports, sufficiently authoritative and based on extensive material. СП were detected only when the E. coli index was \(>1000/100\) [10]. СП concentrations corresponded fully to the sanitary and hygienic conditions at the sampling site, although they were 2–3 orders of magnitude lower than those of coliform bacteria (БГКП), fecal coliforms (ФКП), and enterococci [1]. According to our data, results in habitats up to a certain level of biological pollution, as determined by БГКП and ФКП, were similar to those described above: as the indices of the two indicators rose from 3–3.8 and 2.3–2.6, respectively, the frequency of СП detection increased from 36–48% to 100% of samples, and the НВЧ increased from 2–9 to 48.1–60.5 per liter. Beyond that point, however, although the frequency remained constant at 100%, an increase in the ФКП index to 5.1 per liter and the БГПК index to 5–6 per liter led to a decrease in СП concentrations to 9 and even 2.2 per liter. A similar discrepancy between frequency and density has been described by other researchers [56].

The question of whether the presence of coliforms in water—in practice, fecal pollution—is associated with СП contamination, or whether no such association exists, can be addressed by analyzing studies of water from flowing bodies of water at sites located at varying distances upstream and downstream of wastewater discharges. Indirect evidence of the role of wastewater in СП contamination of water may be provided by the detection of “marked” strains in water receiving wastewater from a hospital where patients infected with carbenicillin-resistant СП were being treated. No СП resistant to this antibiotic were isolated upstream of the wastewater discharge; downstream, they accounted for 18.4% of all strains detected [27]. In a study of a flowing body of water, the НВЧ/100 upstream of the wastewater discharge, 0.5 and 42 km downstream, and in a reservoir was 16.2, 2,025, 223, and 2.3, respectively; the isolation frequencies were 23, 100, 73, and 43.3% (figures are for summer sampling [1]). In a relatively lightly polluted stream in Georgia, USA, only small quantities of СП were detected on average upstream of the wastewater discharge, whereas downstream of the stations the НВЧ/100 ranged from 79 to 230. In a stream more heavily polluted by wastewater discharges, the НВЧ/100 at approximately equal distances between sampling sites in the two streams was 9,500, 1,300, and 750 [33]. In a river receiving wastewater from a medical facility, the mean НВЧ/100 was 470 upstream of the discharge; 5,700 at the discharge point, where mixing of the wastewater with the river water was not yet complete; and 18,000, 9,000, 310, and 33 at distances downstream of 342, 620, 115, and 1,500 m, respectively [27]. Such a rapid decline in СП contamination in surface waters in the United States conflicts with data from microbiologists in Cuba [22] and European countries [1, 17]. Fungicides and herbicides, used in minimal quantities in combination with hydrocarbons, have been found to inhibit СП multiplication [61]. The heavy pollution with these and other preparations used in agriculture may be…

the explanation for the rapid die-off of P. aeruginosa in waters in the United States, although another explanation is possible (see below).

Accounts of the behavior of P. aeruginosa in water differ. It has been reported to multiply in well water [40] and even in river water after purification and disinfection [38]. The introduction of minimal amounts of industrial wastewater into water sharply increases its capacity to multiply [19]. A. Hoadley and his students reported contrary findings [27, 30–32]: in chlorinated water containing no residual chlorine, as in untreated river water, P. aeruginosa dies off within 2–6 hours. Under natural conditions, its density was 70/100 at a point 320 m from a wastewater discharge, but it was not detected 5.1 km downstream. In lake water collected 6 km from a populated area and away from the shore (in stagnant water), P. aeruginosa not only failed to grow but died off within 24 hours [57]. These data suggest that P. aeruginosa is not an autochthonous organism for which water is a permanent habitat [49], but an allochthonous one that periodically enters the aquatic environment and is gradually eliminated from it or multiplies only under favorable conditions.

For practical purposes, when assessing habitats of P. aeruginosa, it is necessary to determine which can serve as a permanent reservoir of Pseudomonas aeruginosa, which can act as a vehicle for infection, and which is merely a temporary habitat [51].

Most reports on testing water for P. aeruginosa do not specify the depths at which samples were collected. Individual studies report sampling depths of 15 cm [15, 57], 15–20 cm [33], 30 cm [42, 54], 50 cm [56], and 60 cm [10, 12]. In the uppermost layer of water (the “microlayer”), where the total bacterial count was \(10^6\), P. aeruginosa accounted for 77% of the bacteria, whereas in a deeper layer the total bacterial count was three orders of magnitude lower and bacteria of the genus Xanthomonas predominated [20]. However, the vigorous motility of P. aeruginosa may overcome its strict aerobic requirements. It has been found at different depths not only in equal or similar quantities—12 per liter at the surface and 2–3 per liter at depth [56], with 35.3% of samples positive at 5 cm and 32.6% at 100 cm below the surface [49]—but also in greater quantities in deeper water layers [43]. A careful analysis accounting for current velocity showed that, in a fast current (1 m/s or more), P. aeruginosa was detected on average in 21.4% of surface samples and in 46.7% of samples at a depth of 1 m; in a slow current (less than 1 m/s), the corresponding figures were 45% and 25%. In bottom sediment, regardless of current velocity, P. aeruginosa was present in 65–72.7% of samples [49]. This led the author to classify it among the epibacteria and to express the belief that P. aeruginosa does not die off in water, as A. Hoadley [27] maintains, but settles to the bottom, carried along by particles in the water.

Various physical factors influence the contamination of water with P. aeruginosa. Depending on meteorological conditions, insolation during water testing reduces the indices of E. coli and acinetobacter but not only has no effect on P. aeruginosa, it even slightly increases its detection [54]. At temperatures below 15 °C, P. aeruginosa was detected in 32% of samples; at 15–30 °C, in 53%; and at 30 °C, in 86% [28]. A large body of data showed a sharp increase in water contamination with P. aeruginosa during the summer [1]. The opposite relationship has also been reported [49].

The behavior of P. aeruginosa in seawater has distinctive features. Although many authors do not identify pseudomonads to the species level, using the general designation Pseudomonas sp. [7, 13, 41, 44, 25, 48], some of their findings can be taken into account. High concentrations of pseudomonads have been reported in seawater. In the uppermost layer, their density reaches \(10^3\)–\(10^4\)/mL in 4 of 5 biotopes examined [7]. In the estuary of the Okhta River (Japan) and Hiroshima Bay, a change in the biological community was observed: enterobacteria and acinetobacter predominated in the estuary, while pseudomonads and vibrios predominated in the bay [25]. In the coastal waters of Nigeria and Ghana and in a lagoon of the Gulf of Guinea, pseudomonads accounted for 9.3% of the total microflora [48]. In lagoon sediment in the waters of the Bahamas, representatives of the genera Vibrio (27.4%) and Pseudomonas (25.5%) predominated [44]. Strictly halophilic species of pseudomonads occur in seawater [13], suggesting that pseudomonads adapted to seawater may exist.

The picture changes sharply when identification is carried to the species level. Thus, in an analysis of freshwater from a 30-acre pond using two comparable methods, the most probable number of P. aeruginosa was 85.5/100 and 179/100, whereas in seawater from a bay with a salinity of 2.1%, the corresponding figures were less than 0.9/100 and 16.7/100 [42]. In Denmark, parallel studies examined water from a freshwater canal running through Copenhagen and seawater from the Øresund strait used for recreation. Both were substantially and comparably polluted by numerous discharges of domestic wastewater. The salinity of the strait did not exceed 0.8–1.0% and varied with wind and current. Of 23 freshwater samples from the canal, 9 (39.1%) contained P. aeruginosa. Of 30 samples from the Øresund taken when wind and current carried wastewater toward the sampling site, 6 (20%) were positive; when wind and current flowed in the opposite direction, only 1 (3.3%) of 30 samples was positive [10]. The comparatively infrequent presence of P. aeruginosa in seawater and the role of fecal pollution in increasing contamination of seawater with this pathogen are beyond doubt [11].

The sanitary and hygienic significance of P. aeruginosa in surface waters has been emphasized by many authors. It has been included as an indicator in official documents in the Federal Republic of Germany [24] and Denmark [10] and was recommended for use in the United States in 1971 [58], although in 1976 its use was restricted to recreational waters [59]. In our country, hygienists and sanitary microbiologists have paid it undeservedly little attention. Only one fundamental study has been published in the domestic literature in recent years [1]. A recently published, scientifically rigorous manual on methods for examining environmental samples [6] does not mention it at all. Yet, given its epidemiological significance as a pathogen, this neglect is unjustified.¹ Water is especially important epidemiologically as a vehicle for its transmission because, unlike enterobacteria, which infect humans only by the oral route, P. aeruginosa has multiple portals of entry. It causes not only diseases of the digestive tract (diarrhea) but also infections of the skin, upper respiratory tract, organs of hearing and vision, and urinary tract. Thus, in water, P. aeruginosa should be regarded both as an indicator and as a pathogen.

References

1. Aleshnya V. V., Tsatka A. A., Vlodavets V. V., Aleshnya E. P. — Gig. i san., 1982, No. 3, pp. 76–77.
2. Vasiliev V. P., Dotsenko A. S. — Zhurn. mikrobiol., 1976, No. 11, pp. 143–144.
3. Kalina G. P. — Ibid., 1984, No. 10, pp. 30–36.
4. Kalina G. P. — Ibid., 1985, No. 5, pp. 91–98.
5. Kornzolova N. B. — Gig. i san., 1983, No. 8, pp. 45–47.
6. Methods for the Sanitary and Microbiological Examination of Environmental Samples / Ed. G. I. Sidorenko. Moscow, 1978.
7. Nizhegorodova L. E., Nidzvetskaya L. M. — Mikrobiol. zhurn., 1980, vol. 42, pp. 143–148.
8. Akock S. — J. Hyg., 1977, vol. 78, pp. 395–409.
9. Blackwall C. et al. — J. med. Microbiol., 1982, vol. 15, pp. 247–251.
10. Bonde G. Bacterial Indicators of Water Pollution. Copenhagen, 1962.
11. Bonde G. — Advanc. in Aquatic Microbiol., 1977, vol. 1, pp. 273–356.
12. Botzenhart K., Röpke S. — Arch. Hyg., 1971, Bd 154, S. 509–516.
13. Brisou J. et al. — Rapp. Comm. int. mer. mediterr., 1976, vol. 23, pp. 133–135.
14. Brodsky M., Ciebin B. — Appl. environm. Microbiol., 1978, vol. 36, pp. 36–42.
15. Buchner G., Stephens J. — Canad. J. Microbiol., 1957, vol. 3, pp. 611–625.
16. Burlingame G. et al. — Appl. environm. Microbiol., 1984, vol. 47, pp. 56–60.
17. Busse M., Seiler H. — In: International Fermentation Symposium. 5th. Berlin, 1976, p. 342.
18. Clark J. — Canad. J. Microbiol., 1969, vol. 15, pp. 721–780.
19. Daubner J. et al. — Zbl. Bakt., I Abt., Orig. B, 1981, Bd 174, S. 453–460.
20. Dott W., Schoenen D. — Zbl. Bakt. I Abt., Orig. B., 1981, Bd 174, S. 174–181.
21. Drake C. — Hlth Lab. Sci., 1966, vol. 3, pp. 10–12.
22. Dutka B., Kwan K. — Appl. environm. Microbiol., 1977, vol. 33, pp. 240–245.
23. Estruch R., Moren R. et al. — Rev. cub. Hig. Epidem., 1982, vol. 20, pp. 613–620.
24. Geuennich H., Müller H. — Zbl. Bakt., I Abt., Orig. A, 1982, Bd 252, S. 230–238.
25. Hashomoto H. — In: Research Related to the UNESCO’s Man and Biosphere Program. Hirosima, 1982–1983, Ser. 1, pp. 32–36.
26. Hoadley A. — J. New Engl. Water Works Ass., 1968, vol. 89, pp. 99–111.
27. Hoadley A. — In: Pseudomonas Aeruginosa, Ecological Aspects and Patient Colonization / Ed. V. M. Young. New York, 1977, pp. 31–57.
28. Hoadley A., Ajello G. — Canad. J. Microbiol., 1972, vol. 18, pp. 1769–1773.
29. Hoadley A., Ajello G., Masterson N. — Appl. Microbiol., 1975, vol. 29, pp. 527–531.
30. Hoadley A., Cheng G. — J. appl. Bact., 1974, vol. 37, pp. 45–57.
31. Hoadley A., MacCoy E. — Appl. Microbiol., 1965, vol. 13, pp. 575–578.
32. Hoadley A., MacCoy E. — Hlth Lab. Sci., 1966, vol. 3, pp. 20–32.
33. Hoadley A., MacCoy E., Rohlich G. — Arch. Hyg., 1968, Bd 282, S. 339–345.
34. Houston A. — In: Royal Committee on the Treating and Disposing of Sewage. 2d. Report. London, 1902, pp. 55–57.
35. Kenner B., Clark H. — J. Water Pollut. Contr. Fed., 1974, vol. 46, pp. 2163–2171.
36. Knorr M., Haas F., Knorr S. — Arch. Hyg., 1956, Bd 140, S. 130–142.
37. v. d. Kooij D. — Anatomie v. Leenwenhoek, 1977, vol. 43, pp. 187–197.
38. v. d. Kooij D. et al. — Appl. environm. Microbiol., 1982, vol. 44, pp. 1086–1095.
39. Lanyi B., Gregacs M., Adam M. — Acta microbiol. Acad. Sci. hung., 1966, vol. 13, pp. 319–326.
40. Laurence R. et al. — Ann. Microbiol., 1981, vol. 132-B, pp. 455–463.
41. Leifson E. — Appl. Microbiol., 1968, vol. 16, pp. 1633–1634.
42. Levin M., Cabelli V. — Ibid., 1972, vol. 24, pp. 864–870.
43. Lighthart B. — Canad. J. Microbiol., 1975, vol. 21, pp. 392–394.
44. MacCallum M. — J. appl. Bact., 1970, vol. 33, pp. 533–542.
45. Némedi L., Lanyi B. — Egészségtudomány, 1970, vol. 14, pp. 334–341.
46. Neu H. — In: Glucose Nonfermenting, Gram-Negative Bacteria in Clinical Microbiology / Ed. G. Gilardi. West Palm Beach, 1978, pp. 83–104.
47. Ortiz J. et al. — Conference Scientific of National Parks. 1st. Abstract. Amherst. Mass, 1976, p. 57.
48. Otunola E. et al. — Microbios Lett., 1983, vol. 24, pp. 91–98.
49. Pellett S. et al. — Appl. environm. Microbiol., 1983, vol. 45, pp. 328–332.
50. Reinhardt D. et al. — In: Developments in Industrial Microbiology. Houston, 1979, pp. 705–721.
51. Rhame F. — In: Pseudomonas Aeruginosa, the Organism, Diseases it Causes, and Their Treatment / Ed. L. Sabath. Bern, 1980, pp. 31–51.
52. Ringen L., Drake Ch. — J. Bact., 1952, vol. 64, pp. 841–845.
53. Schmidt-Lorenz W. — Brauerei Rundsch., 1983, Bd 94, S. 62–72.
54. Seifried P. — In: Conference on Great Lakes Research. 10th. Proceedings. Ann Harbour, 1973, pp. 163–182.
55. Selenka F. — Arch. Hyg., 1960, Bd 144, S. 627–634.

¹ Only in 1984 did the USSR Ministry of Health approve methodological recommendations for the detection and identification of Pseudomonas aeruginosa in environmental samples, including water.

1. Selenka F., Ruschke R. — Ibid., 1965, Bd. 149, S. 273–287.
2. Sjogren K., Gibson M. — Appl. environm. Microbiol., 1981, vol. 41, p. 1331–1336.
3. Standard Methods for the Examination of Water and Wastewater (APHA, USA), 13th Ed. Washington, 1971.
4. Stabdard Methods for the Examination of Water and Wastwater (APHA, USA), 14th Ed. Washington, 1976.
5. Thresh I. C. The Examination of Water and Water Supplies. London, 1958.
6. Toth D. et al. — Biologia (Bratisl.), 1981, vol. 36, p. 775–783.

Received 25 March 1985.