APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Nov. 1990, p. 3463-3467
Vol. 56, No. 11
0099-2240/90/113463-05$02.00/0 Copyright © 1990, American Society for Microbiology
Survival of Bacteria during Aerosolization B. MARTHI,1* V. P. FIELAND,' M. WALTER,' AND R. J. SEIDLER2 NSI Technology Services Corp.' and U.S. Environmental Protection Agency Environmental Research Laboratory,2 Corvallis, Oregon 97333 Received 23 March 1990/Accepted 17 August 1990
One form of commercial application of microorganisms, including genetically engineered microorganisms is as an aerosol. To study the effect of aerosol-induced stress on bacterial survival, nonrecombinant spontaneous antibiotic-resistant mutants of four organisms, Enterobacter cloacae, Erwinia herbicola, Klebsiella planticola, and Pseudomonas syringae, were sprayed in separate experiments in a greenhouse. Samples were collected over a distance of 15 m from the spray site for enumeration. Spores of Bacillus subtilis were used as tracers to estimate the effects of dilution on changes in population over distance. Viable counts of P. syringae, Enterobacter cloacae, and K. planticola decreased significantly over a distance of 15 m. Erwinia herbicola showed no significant decline in counts over the same distance. The degree of survival of P. syringae during aerosolization was dependent on ambient environmental conditions (i.e., temperature, relative humidity), droplet size of the aerosol, and prior preparative conditions. Survival was greatest at high relative humidities (70 to 80%) and low temperatures (12°C). Survival was reduced when small droplet sizes were used. The process of washing the cells prior to aerosolization also caused a reduction in their survival. Results from these experiments will be useful in developing sound methodologies to optimize enumeration and for predicting the downwind dispersal of airborne microorganisms, including genetically engineered microorganisms. In recent years, there has been an increased awareness of the environmental applications of genetically engineered microorganisms. The impact of this technology will probably be felt in the field of agricultural biotechnology, most notably, in the areas of microbial control of plant pathogens, weeds, and insects (3, 15), and also in the area of bioremediation. The release of genetically engineered microorganisms into natural environments has raised the issue of proper monitoring and control of these organisms, from a standpoint of both human health and protection of the environment. Past research on aerosolized bacteria has been done under very controlled conditions in contained laboratory chambers. Data from such studies indicate that aerosolization can cause stress on bacterial populations (2, 4-6, 18). The survival of aerosolized bacteria is influenced by (i) growth conditions prior to aerosolization; (ii) environmental conditions during aerosolization; (iii) methods of aerosolization; and (iv) methods of collection and enumeration (2-8, 12-14). Temperature and relative humidity during aerosolization have been shown to influence the survival of aerosolized bacteria. Escherichia coli K-12 was shown to survive better at low humidities and temperatures than at high humidities when sprayed into an atmosphere of nitrogen (5). Survival, however, was reduced even at low humidities in an atmosphere of oxygen (4). Survival is also dependent on the species of organism being aerosolized. The only extensive field release study was the release of an ice-nucleation-active strain of Pseudomonas syringae (16). That study indicated that aerosolization may affect the enumeration, survival, and function of bacterial populations and has also raised concerns regarding the survival and efficient detection of genetically engineered microorganisms released into the environment. It is expected that, in many instances, future releases of microorganisms, including genetically engineered microorganisms, will be as aerosols. *
Therefore, it becomes necessary to study the effects of aerosolization on the survival of microorganisms. This paper presents the results of an investigation into the effects of aerosolization on the survival of vegetative bacteria and the influence of ambient environmental and culture preparative conditions on their enumeration. Spores of Bacillus subtilis were used as physical tracers to determine the extent of dilution of the aerosolized population, since their survival is not affected by most adverse conditions (12, 17). Four different bacterial genera were tested to evaluate variations in their survival during aerosolization. Results indicate that temperature, relative humidity, droplet size, and washing of cells prior to aerosolization affect the survival of aerosolized bacteria. MATERIALS AND METHODS used. Organisms Target organisms used in this study were Enterobacter cloacae EPA 117 (isolated from the gut of a cutworm larva; resistant to 500 ,ug of nalidixic acid per ml), Erwinia herbicola EPA 81A (supplied by Steve Beer, Cornell University, Ithaca, N.Y.; resistant to 500 ,ug of nalidixic acid per ml), Klebsiella planticola EPA 658 (isolated from roots of a radish plant; resistant to 100 ,ug of rifampin per ml), and P. syringae TLP2 (supplied by Steve Lindow, University of California, Berkeley; resistant to 100 ,ug of rifampin per ml). Spores of B. subtilis ATCC 6537 were used as internal controls to estimate the extent of dilution of the aerosolized
population. Preparation of target cell suspensions. Target organisms were grown at 30°C with shaking at 150 rpm in 600 ml of Luria-Bertani broth (LB broth; Difco Laboratories, Detroit, Mich.) supplemented with the appropriate antibiotic. Antibiotics were used to suppress the growth of indigenous airborne microorganisms and to allow for the selective enumeration of the test microorganism. The organisms were grown to an A6m0 of 2.0, corresponding to a viable cell density of approximately 109 cells per ml. Cells were washed three times in sterile 10 mM phosphate buffer (per liter of distilled water: 4.0 g of KH2PO4, 13.6 g of K2HPO4; pH 7.2)
Corresponding author. 3463
APPL. ENVIRON. MICROBIOL.
MARTHI ET AL.
3464
Meteoroloial Station
Spray
0-
Source
1-
3t--
*-C--
5t
0.@..
Wind Direction
a)
10 _
*
00
15 FIG. 1. Experimental design. Filled circles represent location of AGIs.
and suspended in 300 ml of sterile distilled water. The washed cell suspension was then added to 2.7 liters of sterile distilled water in a stainless-steel CO2 sprayer (R & D Sprayers, Inc., Opelousas, La.). For experiments with unwashed cells, the bacterial suspensions were centrifuged and suspended directly in sterile distilled water prior to spraying. Preparation of B. subtilis spore suspension. B. subtilis cells were grown on plates of nutrient agar (Difco) containing 2.5 ppm (2.5 jig/ml) of MnSO4 for 7 to 10 days at 30°C to induce sporulation. The spores were removed from the plates by suspension in distilled water (10 ml per plate). The spore suspension was then washed three times and suspended in the same volume of sterile distilled water. Prior to spraying, 75 to 100 ml of the washed spore suspension was added to the spray suspension. Aerosolization procedure. Aerosolization experiments were performed (i) to compare survival of Enterobacter cloacae, Erwinia herbicola, K. planticola, and P. syringae; (ii) to study the effects of droplet sizes on survival; and (ii) to study the effects of ambient temperature and relative humidity on survival. All studies on the effects of various environmental and preparative conditions on survival were carried out with aerosolized P. syringae. All experiments were conducted in a greenhouse (30 by 9 m). All-glass impinger air samplers (AGIs; Ace Glass Inc., Vineland, N.J.) were located at distances of 1, 3, 5, 10, and 15 m from the spray site (Fig. 1). Each AGI contained 20 ml of sterile 10 mM phosphate buffer. The flow rate of air through the critical orifice of the AGIs was 12.5 liters/min. A nonaerosolized control was run upwind of the release site by adding 1 ml of the spray suspension to 20 ml of buffer in an AGI. The cell suspension was sprayed at a pressure of 35 lb/in2 for 2 min. Two different particle sizes were released in separate experiments by using different spray nozzles: (i) Teejet 8004-SS, delivering a median diameter of 450 p.m
(Spraying Systems Co., Wheaton, Ill.); (ii) Delavan HC 2-70°, delivering a median diameter of 130 p.m (Delavan, Inc., Lexington, Tenn.). The AGIs were run during the spray and for a period of 18 min following the spray. Ambient environmental conditions (temperature, relative humidity, wind speed, and direction) were monitored with various meteorological equipment and stored in a 21XL Datalogger (Campbell Scientific, Logan, Utah). The temperature and relative humidity were adjusted as appropriate for each experiment. The wind speed was approximately 1 m/s and did not vary significantly between experiments. The wind speed was controlled by two large fans at one end of the greenhouse. The sampling setup was adjusted in such a way that the bacterial spray travelled in the direction of the wind, and the AGIs were located directly downstream from the spray source (Fig. 1). Enumeration of cells. Following aerosolization, aliquots of samples were first plated on LB agar containing the appropriate antibiotics for the enumeration of target organisms. Cycloheximide was added at a concentration of 100 ,ug/ml to all enumeration media to inhibit fungal growth. To enumerate B. subtilis spores, samples were heated for 20 min at 80°C and plated on nutrient agar plates. Plates were incubated at 30°C for 36 to 48 h. Analysis of data. For each of the experiments described here, data were collected from three separate experiments with six replicates of each. Spores were used as internal controls to determine the extent of dilution of the population. The viable counts of spores were compared with those of target organisms to determine the effects of aerosolization and environmental stresses on the survival of target microorganisms. All data are expressed as target/spore ratios. Spore counts were normalized to target cell counts prior to calculating target/spore ratios. Target/spore ratios were calculated by dividing the target cell counts by the normalized spore counts. A change in the target/spore ratio indicates that there is a differential response of the target organism over and above the dilution effect. Thus, a reduction in the survival of the target organism is indicated by a reduction in the target/spore ratio. An analysis of variance was performed on target/spore ratios, using the Statistical Analysis System (SAS Institute, Inc., Cary, N.C.). Ratios were compared between each replicate to determine whether there were any significant variations between replicates. Target/spore ratios were also compared at each of the distances to determine how survival was affected by distance travelled. P values of 0.05 or lower were considered significant. All data points in the figures are represented by the mean target/spore ratio of all replicates ± 1 standard error.
RESULTS The effect of the aerosolization process on the survival of aerosolized Enterobacter cloacae, Erwinia herbicola, K. planticola, and P. syringae at 22°C and a relative humidity of 39% is shown in Fig. 2. In this experiment, and all subsequent experiments, the test organism was sprayed out at an initial concentration of ca. 108 CFU/ml. There was a reduction of about 4 orders of magnitude in the numbers of cells recovered at the farthest distance (15 m) from the spray site. Erwinia herbicola showed no significant reduction in target/ spore ratios over the 15-m sampling distance, indicating that there was negligible reduction in the survival of this organism compared with that of the spores. The survival of the other three organisms was significantly lower than that of Erwinia herbicola under the same test conditions. The
VOL. 56, 1990
SURVIVAL OF BACTERIA DURING AEROSOLIZATION
3.0 0
E. E K P
25.-
4.5
herbicola 0 0 cloacce 0*0 A pianticolc A A syringce Av
,
ol
L.Li 1-Y
-D
a-
(,n
1.5-
L'i
.-D ll.
Lii :7.
-
(a)
120C 0 22°C 0
4 04.0 3.5 -
0 *
ar
Cl.
1-1
3465
1.
0+
O. 5,
0
0
iEVi IiEfZZT
a-
2.5-
if)
,2-
- * j~~~~~~~~~~~
A *
3.0-
-
202 01.5 -
z
1.0-
>
050.0-
0.0 10 5 ~10 SAMPLING DISTANCE (Meters)
S.
r
.L'
cc0
survival patterns of Enterobacter cloacae, K. planticola, and P. syringae were not significantly different from each other under the conditions tested. At a distance of 1 m, organisms sprayed by using a large droplet size (median diameter, 450 ,um) demonstrated a significantly higher degree of survival than organisms sprayed by using a smaller droplet size (median diameter, 130 pm) at a temperature of 22°C and relative humidity of 39% (Fig. 3). At distances beyond 1 m, there was no significant reduction in the target/spore ratios between the two droplet sizes. Survival at 10 and 15 m was significantly lower when compared with survival at 1 and 3 m for both droplet sizes as indicated by a reduction in the target/spore ratios. When sprayed as large droplets, there was no significant reduction in the ratios, indicating that survival was higher at an ambient spray temperature of 12°C compared with an
0 0
+ -0
10 0 5
10
15
20
SAMPLING DISTANCE (Meters)
FIG. 3. Effect of droplet size on survival of aerosolized P. syringae. P. syringae was aerosolized at 22°C and 39% relative humidity as described in the text. Samples were collected and enumerated as described in the text. Large droplet size, Median droplet diameter of 450 p.m. Small droplet size, Median droplet diameter of 130 p.m. Each data point represents the mean + 1 standard error of 12 replicates. Absence of error bars indicates that the amount of error was too small to fit in the scale of this graph.
(b)
120CO
I22°C
,
.-
0
0
*
1~ ~ ~ ~
60
ar,
0.5+
20
15
0
0~~~~
Li
1.0
10
SAMPLING DISTANCE (Meters)
FIG. 2. Comparative survival of bacteria in aerosols. Target organisms were grown in LB broth containing the appropriate antibiotics, washed, and sprayed at a concentration of approximately 107 cells per ml, as described in the text. Aerosol samples were collected at distances of 1, 3, 5, 10, and 15 m and enumerated as described in the text. Each data point represents the mean + 1 standard error of six replicates.
Large Droplet Size 0 Small Droplet Size 0
5
0
20
4.0
1-I
z ,
I0 0
5
10
15
20
SAMPLING DISTANCE (Meters)
FIG. 4. Effect of ambient temperature on survival of aerosolized
P. syringae. P. syringae was aerosolized as described in the text. Samples were collected and enumerated as described in the text. (a) Large droplet size, Median droplet diameter of 450 p.m. (b) Small droplet size, Median droplet diameter of 130 p.m. Each data point represents the mean + 1 standard error of 12 replicates. Absence of error bars indicates that the amount of error was too small to fit in the scale of this graph.
ambient spray temperature of 22°C (Fig. 4a). The relative survival at the lower temperature (12°C) was greatly enhanced when a smaller droplet size was used (Fig. 4b). The experiments described above were done with suspensions of P. syringae which were washed three times in sterile 10 mM phosphate buffer prior to spraying. Unwashed organisms sprayed as large droplets showed no significant decrease in numbers over the entire 15-m sampling distance. However, unwashed organisms sprayed as small droplets showed significantly reduced target/spore ratios at distances of 5, 10, and 15 m (Fig. 5).
DISCUSSION Experiments with four different gram-negative bacteria (Enterobacter cloacae, Erwinia herbicola, K. planticola, and P. syringae) demonstrated that, under similar release conditions (22°C and 39% relative humidity), the different organisms showed different degrees of survival during aerosolization. Erwinia herbicola showed the highest level of survival at ambient temperatures of 22°C (Fig. 2). There were no significant differences in the survival of the other three organisms. Under our test conditions, the detrimental effects of aerosolization on the survival of washed P. syringae were
APPL. ENVIRON. MICROBIOL.
MARTHI ET AL.
3466
3.0 0
Large Droplet Size 0 Small Droplet Size 0
2.5
0 0
0 Q~
2.0 -\
.L
0.
LL-
1.5 0-
0
1.0* O
z
L'-
O
0.5
0'.00
t 5
10
15
2(10
SAMPLING DISTANCE (Meters)
FIG. 5. Survival of unwashed P. syringae during aerosolization. P. syringae was aerosolized at 22°C and 39% RH as described in the text. Droplet sizes are the same as in the legends to Fig. 3 and 4. Each data point represents the mean + 1 standard error of 12 replicates. Absence of error bars indicates that the amount of error was too small to fit in the scale of this graph.
independent of the size of the droplets carrying the bacteria at a higher ambient temperature (22°C; Fig. 4). At an ambient temperature of 22°C and a relative humidity of 39%, the rate of evaporation of droplets is very rapid, even for large droplets (10 to 30 sec; W. Yates, personal communication). Upon evaporation, bacterium-carrying droplets reach a critical final size beyond which there is no further reduction in size. This critical size is dependent on ambient environmental conditions (B. Lighthart, B. T. Shaffer, B. Marthi, and L. Ganio, unpublished data). The absence of any significant differences in the survival of P. syringae between the two droplet sizes at 22°C is probably indicative of the fact that the final particle size in both cases is similar. The rate of evaporation, which is a function of ambient temperature and relative humidity, is reduced at the lower temperature (12°C), since the relative humidity is high (77% under the conditions tested). A relative humidity in the range of 70 to 80% has been shown to have a protective effect on aerosolized bacteria (8, 11). The protective effect seen in Fig. 4 may, therefore, be a combined effect of lower temperature and higher relative humidity. When experiments were conducted with small droplet sizes at an ambient temperature of 12°C, results showed a higher relative target/spore ratio at distances of 10 and 15 m than at distances of 3 and 5 m. It is thought that this may be due to clumping together of B. subtilis spores, used as controls in all experiments, due to the effects of the increased humidity. This may underestimate the number of spores counted and also the relative ratio of target organisms to spores. A high relative humidity may also cause clumping together of vegetative cells and possibly increase survival.
Preparative procedures prior to aerosolization had a profound effect on the survival of P. syringae (Fig. 5). These experiments were done under the same conditions of temperature and relative humidity (22°C and 39%) as the experiments described in the legend to Fig. 2. Unwashed cells sprayed in large droplets did not show any significant decrease in survival over a distance of 15 m (Fig. 5). Washed cells aerosolized under the same conditions showed a significant reduction in survival (Fig. 3). Unwashed cells sprayed as small droplets, however, showed significantly reduced survival. The presence of organic matter from the growth medium (LB broth) increases the vapor pressure of droplets.
This might reduce evaporation and confer protection to the aerosolized bacteria. However, this protective effect may be masked by the more rapid evaporation of smaller droplets, thus reflecting the greater die-off seen in Fig. 5. Due to the short (20-min) collection time in the AGIs and the absence of any nutrient medium (cells are collected in phosphate buffer), the possibility that cells may be multiplying is remote. Additional experiments were done to determine whether the washing procedure caused any reduction in the viable counts of these bacteria. Viable counts of an unwashed suspension of P. syringae were compared with viable counts of the organism washed three times with 10 mM phosphate buffer (pH 7.2). Results showed that there was no significant difference in the viable counts of unwashed and washed cells (data not shown). This indicates that the washing procedure itself did not affect viable counts of these bacteria. Thus, any differences seen in the survival of washed and unwashed bacteria after aerosolization were due to the stress of the aerosolization process itself. In summary, of the four genera used in this study, Erwinia herbicola showed the highest degree of survival during aerosolization over distances of 15 m. There was no significant difference in the survival of Enterobacter cloacae, K. planticola, and P. syringae. The survival of aerosolized P. syringae was found to be dependent on a number of factors, including droplet size, ambient temperature, and relative humidity, and culture preparative procedures. This study is useful for the development of monitoring protocols for detection of released microorganisms, including genetically engineered microorganisms. Such protocols would include the use of media and conditions that would better allow stressed microorganisms to be detected. Results from such studies can also be used to model the downwind dispersal of aerosolized bacteria by predicting the effects of various environmental and preparative conditions on the survival of aerosolized bacteria. ACKNOWLEDGMENTS We are grateful to David Zirkle and Lisa Ganio, NSI Technology Services Corp., for help in the statistical analysis of data presented in this manuscript. LITERATURE CITED 1. Beebe, J. M., and G. B. Pirsch. 1958. Response of air-borne species of Pasteurella to artificial radiation simulating sunlight under different conditions of relative humidity. Appl. Microbiol. 6:127-138. 2. Benbough, J. E. 1967. Death mechanisms in airborne Escherichia coli. J. Gen. Microbiol. 47:325-333. 3. Blakeman, J. P., and N. J. Fokkema. 1982. Potential for biological control of plant diseases on the phylloplane. Annu. Rev. Phytopathol. 20:167-192. 4. Cox, C. S. 1966. The survival of Escherichia coli sprayed into air and into nitrogen from distilled water and from solutions of protecting agents, as a function of relative humidity. J. Gen. Microbiol. 44:15-31. 5. Cox, C. S. 1968. The aerosol survival and cause of death of Escherichia coli K12. J. Gen. Microbiol. 54:169-175. 6. Cox, C. S. 1969. The cause of loss of viability of airborne Escherichia coli K12. J. Gen. Microbiol. 57:77-80. 7. Cox, C. S. 1971. Aerosol survival of Pasteurella tularensis disseminated from the wet and dry states. Appl. Microbiol. 21:482-486. 8. Cox, C. S. 1987. Relative humidity and temperature, p. 172-205. In The aerobiological pathway of microorganisms. John Wiley & Sons, Inc., New York. 9. Cox, C. S., and F. Baldwin. 1967. The toxic effect of oxygen
VOL. 56, 1990
the aerosol survival of Escherichia coli B. J. Gen. Microbiol. 49:115-117. Dorsey, E. L., R. F. Berendt, and E. L. Neff, Jr. 1970. Effect of sodium fluorescein and plating medium on recovery of irradiated Escherichia coli and Serratia marcescens from aerosols. Appl. Microbiol. 20:834-838. Dunklin, E. W., and T. T. Puck. 1948. The lethal effect of relative humidity on air-borne bacteria. J. Exp. Med. 87:87-101. Ehrlich, R., S. Miller, and R. L. Walker. 1970. Relationship between atmospheric temperature and survival of airborne bacteria. Appl. Microbiol. 19:245-249. Hatch, M. T., and R. L. Dimmick. 1966. Physiological response of airborne bacteria to shifts in relative humidity. Bacteriol. Rev. 30:597-603. Hess, G. E. 1965. Effects of oxygen on aerosolized Serratia marcescens. Appl. Microbiol. 13:781-787. Lindow, S. E. 1985. Ecology of Pseudomonas syringae relevant to the field use of Ice-deletion mutants constructed in vitro for plant frost control, p. 23-35. In H. 0. Halvorson, D. Pramer, and M. Rogul (ed.), Engineered organisms in the environment: scientific issues. American Society for Microbiology, Washington, D.C. Lindow, S. E., G. Knudsen, R. J. Seidler, M. V. Walter, V. M.
upon
10.
11.
12. 13. 14.
15.
16.
SURVIVAL OF BACTERIA DURING AEROSOLIZATION
3467
Lambou, P. S. Amy, D. Schmedding, V. Prince, and S. Hern. 1988. Aerial dispersal and epiphytic survival of Pseudomonas syringae during a pretest for the release of genetically engineered strains into the environment. Appl. Environ. Microbiol.
54:1557-1563. 17. Miller, W. S., R. A. Scherff, C. R. Piepoli, and L. S. Idoine. 1961. Physical tracers for bacterial aerosols. Appl. Microbiol. 9:248-251. 18. Poon, C. P. C. 1966. Studies on the instantaneous death of airborne Escherichia coli. Am. J. Epidemiol. 84:1-9. 19. Webb, S. J. 1959. Factors affecting the viability of air-borne bacteria. I. Bacteria aerosolized from distilled water. Can. J. Microbiol. 5:649-669. 20. Webb, S. J. 1960. Factors affecting the viability of air-borne bacteria. II. The effect of chemical additives on the behavior of air-borne cells. Can. J. Microbiol. 6:71-87. 21. Webb, S. J. 1960. Factors affecting the viability of air-borne bacteria. III. The role of bonded water and protein structure in the death of air-borne cells. Can. J. Microbiol. 6:89-105. 22. Webb, S. J. 1967. The influence of oxygen and inositol on the survival of semidried microorganisms. Can. J. Microbiol. 13: 733-742.
Vol. 56, No. 11
0099-2240/90/113463-05$02.00/0 Copyright © 1990, American Society for Microbiology
Survival of Bacteria during Aerosolization B. MARTHI,1* V. P. FIELAND,' M. WALTER,' AND R. J. SEIDLER2 NSI Technology Services Corp.' and U.S. Environmental Protection Agency Environmental Research Laboratory,2 Corvallis, Oregon 97333 Received 23 March 1990/Accepted 17 August 1990
One form of commercial application of microorganisms, including genetically engineered microorganisms is as an aerosol. To study the effect of aerosol-induced stress on bacterial survival, nonrecombinant spontaneous antibiotic-resistant mutants of four organisms, Enterobacter cloacae, Erwinia herbicola, Klebsiella planticola, and Pseudomonas syringae, were sprayed in separate experiments in a greenhouse. Samples were collected over a distance of 15 m from the spray site for enumeration. Spores of Bacillus subtilis were used as tracers to estimate the effects of dilution on changes in population over distance. Viable counts of P. syringae, Enterobacter cloacae, and K. planticola decreased significantly over a distance of 15 m. Erwinia herbicola showed no significant decline in counts over the same distance. The degree of survival of P. syringae during aerosolization was dependent on ambient environmental conditions (i.e., temperature, relative humidity), droplet size of the aerosol, and prior preparative conditions. Survival was greatest at high relative humidities (70 to 80%) and low temperatures (12°C). Survival was reduced when small droplet sizes were used. The process of washing the cells prior to aerosolization also caused a reduction in their survival. Results from these experiments will be useful in developing sound methodologies to optimize enumeration and for predicting the downwind dispersal of airborne microorganisms, including genetically engineered microorganisms. In recent years, there has been an increased awareness of the environmental applications of genetically engineered microorganisms. The impact of this technology will probably be felt in the field of agricultural biotechnology, most notably, in the areas of microbial control of plant pathogens, weeds, and insects (3, 15), and also in the area of bioremediation. The release of genetically engineered microorganisms into natural environments has raised the issue of proper monitoring and control of these organisms, from a standpoint of both human health and protection of the environment. Past research on aerosolized bacteria has been done under very controlled conditions in contained laboratory chambers. Data from such studies indicate that aerosolization can cause stress on bacterial populations (2, 4-6, 18). The survival of aerosolized bacteria is influenced by (i) growth conditions prior to aerosolization; (ii) environmental conditions during aerosolization; (iii) methods of aerosolization; and (iv) methods of collection and enumeration (2-8, 12-14). Temperature and relative humidity during aerosolization have been shown to influence the survival of aerosolized bacteria. Escherichia coli K-12 was shown to survive better at low humidities and temperatures than at high humidities when sprayed into an atmosphere of nitrogen (5). Survival, however, was reduced even at low humidities in an atmosphere of oxygen (4). Survival is also dependent on the species of organism being aerosolized. The only extensive field release study was the release of an ice-nucleation-active strain of Pseudomonas syringae (16). That study indicated that aerosolization may affect the enumeration, survival, and function of bacterial populations and has also raised concerns regarding the survival and efficient detection of genetically engineered microorganisms released into the environment. It is expected that, in many instances, future releases of microorganisms, including genetically engineered microorganisms, will be as aerosols. *
Therefore, it becomes necessary to study the effects of aerosolization on the survival of microorganisms. This paper presents the results of an investigation into the effects of aerosolization on the survival of vegetative bacteria and the influence of ambient environmental and culture preparative conditions on their enumeration. Spores of Bacillus subtilis were used as physical tracers to determine the extent of dilution of the aerosolized population, since their survival is not affected by most adverse conditions (12, 17). Four different bacterial genera were tested to evaluate variations in their survival during aerosolization. Results indicate that temperature, relative humidity, droplet size, and washing of cells prior to aerosolization affect the survival of aerosolized bacteria. MATERIALS AND METHODS used. Organisms Target organisms used in this study were Enterobacter cloacae EPA 117 (isolated from the gut of a cutworm larva; resistant to 500 ,ug of nalidixic acid per ml), Erwinia herbicola EPA 81A (supplied by Steve Beer, Cornell University, Ithaca, N.Y.; resistant to 500 ,ug of nalidixic acid per ml), Klebsiella planticola EPA 658 (isolated from roots of a radish plant; resistant to 100 ,ug of rifampin per ml), and P. syringae TLP2 (supplied by Steve Lindow, University of California, Berkeley; resistant to 100 ,ug of rifampin per ml). Spores of B. subtilis ATCC 6537 were used as internal controls to estimate the extent of dilution of the aerosolized
population. Preparation of target cell suspensions. Target organisms were grown at 30°C with shaking at 150 rpm in 600 ml of Luria-Bertani broth (LB broth; Difco Laboratories, Detroit, Mich.) supplemented with the appropriate antibiotic. Antibiotics were used to suppress the growth of indigenous airborne microorganisms and to allow for the selective enumeration of the test microorganism. The organisms were grown to an A6m0 of 2.0, corresponding to a viable cell density of approximately 109 cells per ml. Cells were washed three times in sterile 10 mM phosphate buffer (per liter of distilled water: 4.0 g of KH2PO4, 13.6 g of K2HPO4; pH 7.2)
Corresponding author. 3463
APPL. ENVIRON. MICROBIOL.
MARTHI ET AL.
3464
Meteoroloial Station
Spray
0-
Source
1-
3t--
*-C--
5t
0.@..
Wind Direction
a)
10 _
*
00
15 FIG. 1. Experimental design. Filled circles represent location of AGIs.
and suspended in 300 ml of sterile distilled water. The washed cell suspension was then added to 2.7 liters of sterile distilled water in a stainless-steel CO2 sprayer (R & D Sprayers, Inc., Opelousas, La.). For experiments with unwashed cells, the bacterial suspensions were centrifuged and suspended directly in sterile distilled water prior to spraying. Preparation of B. subtilis spore suspension. B. subtilis cells were grown on plates of nutrient agar (Difco) containing 2.5 ppm (2.5 jig/ml) of MnSO4 for 7 to 10 days at 30°C to induce sporulation. The spores were removed from the plates by suspension in distilled water (10 ml per plate). The spore suspension was then washed three times and suspended in the same volume of sterile distilled water. Prior to spraying, 75 to 100 ml of the washed spore suspension was added to the spray suspension. Aerosolization procedure. Aerosolization experiments were performed (i) to compare survival of Enterobacter cloacae, Erwinia herbicola, K. planticola, and P. syringae; (ii) to study the effects of droplet sizes on survival; and (ii) to study the effects of ambient temperature and relative humidity on survival. All studies on the effects of various environmental and preparative conditions on survival were carried out with aerosolized P. syringae. All experiments were conducted in a greenhouse (30 by 9 m). All-glass impinger air samplers (AGIs; Ace Glass Inc., Vineland, N.J.) were located at distances of 1, 3, 5, 10, and 15 m from the spray site (Fig. 1). Each AGI contained 20 ml of sterile 10 mM phosphate buffer. The flow rate of air through the critical orifice of the AGIs was 12.5 liters/min. A nonaerosolized control was run upwind of the release site by adding 1 ml of the spray suspension to 20 ml of buffer in an AGI. The cell suspension was sprayed at a pressure of 35 lb/in2 for 2 min. Two different particle sizes were released in separate experiments by using different spray nozzles: (i) Teejet 8004-SS, delivering a median diameter of 450 p.m
(Spraying Systems Co., Wheaton, Ill.); (ii) Delavan HC 2-70°, delivering a median diameter of 130 p.m (Delavan, Inc., Lexington, Tenn.). The AGIs were run during the spray and for a period of 18 min following the spray. Ambient environmental conditions (temperature, relative humidity, wind speed, and direction) were monitored with various meteorological equipment and stored in a 21XL Datalogger (Campbell Scientific, Logan, Utah). The temperature and relative humidity were adjusted as appropriate for each experiment. The wind speed was approximately 1 m/s and did not vary significantly between experiments. The wind speed was controlled by two large fans at one end of the greenhouse. The sampling setup was adjusted in such a way that the bacterial spray travelled in the direction of the wind, and the AGIs were located directly downstream from the spray source (Fig. 1). Enumeration of cells. Following aerosolization, aliquots of samples were first plated on LB agar containing the appropriate antibiotics for the enumeration of target organisms. Cycloheximide was added at a concentration of 100 ,ug/ml to all enumeration media to inhibit fungal growth. To enumerate B. subtilis spores, samples were heated for 20 min at 80°C and plated on nutrient agar plates. Plates were incubated at 30°C for 36 to 48 h. Analysis of data. For each of the experiments described here, data were collected from three separate experiments with six replicates of each. Spores were used as internal controls to determine the extent of dilution of the population. The viable counts of spores were compared with those of target organisms to determine the effects of aerosolization and environmental stresses on the survival of target microorganisms. All data are expressed as target/spore ratios. Spore counts were normalized to target cell counts prior to calculating target/spore ratios. Target/spore ratios were calculated by dividing the target cell counts by the normalized spore counts. A change in the target/spore ratio indicates that there is a differential response of the target organism over and above the dilution effect. Thus, a reduction in the survival of the target organism is indicated by a reduction in the target/spore ratio. An analysis of variance was performed on target/spore ratios, using the Statistical Analysis System (SAS Institute, Inc., Cary, N.C.). Ratios were compared between each replicate to determine whether there were any significant variations between replicates. Target/spore ratios were also compared at each of the distances to determine how survival was affected by distance travelled. P values of 0.05 or lower were considered significant. All data points in the figures are represented by the mean target/spore ratio of all replicates ± 1 standard error.
RESULTS The effect of the aerosolization process on the survival of aerosolized Enterobacter cloacae, Erwinia herbicola, K. planticola, and P. syringae at 22°C and a relative humidity of 39% is shown in Fig. 2. In this experiment, and all subsequent experiments, the test organism was sprayed out at an initial concentration of ca. 108 CFU/ml. There was a reduction of about 4 orders of magnitude in the numbers of cells recovered at the farthest distance (15 m) from the spray site. Erwinia herbicola showed no significant reduction in target/ spore ratios over the 15-m sampling distance, indicating that there was negligible reduction in the survival of this organism compared with that of the spores. The survival of the other three organisms was significantly lower than that of Erwinia herbicola under the same test conditions. The
VOL. 56, 1990
SURVIVAL OF BACTERIA DURING AEROSOLIZATION
3.0 0
E. E K P
25.-
4.5
herbicola 0 0 cloacce 0*0 A pianticolc A A syringce Av
,
ol
L.Li 1-Y
-D
a-
(,n
1.5-
L'i
.-D ll.
Lii :7.
-
(a)
120C 0 22°C 0
4 04.0 3.5 -
0 *
ar
Cl.
1-1
3465
1.
0+
O. 5,
0
0
iEVi IiEfZZT
a-
2.5-
if)
,2-
- * j~~~~~~~~~~~
A *
3.0-
-
202 01.5 -
z
1.0-
>
050.0-
0.0 10 5 ~10 SAMPLING DISTANCE (Meters)
S.
r
.L'
cc0
survival patterns of Enterobacter cloacae, K. planticola, and P. syringae were not significantly different from each other under the conditions tested. At a distance of 1 m, organisms sprayed by using a large droplet size (median diameter, 450 ,um) demonstrated a significantly higher degree of survival than organisms sprayed by using a smaller droplet size (median diameter, 130 pm) at a temperature of 22°C and relative humidity of 39% (Fig. 3). At distances beyond 1 m, there was no significant reduction in the target/spore ratios between the two droplet sizes. Survival at 10 and 15 m was significantly lower when compared with survival at 1 and 3 m for both droplet sizes as indicated by a reduction in the target/spore ratios. When sprayed as large droplets, there was no significant reduction in the ratios, indicating that survival was higher at an ambient spray temperature of 12°C compared with an
0 0
+ -0
10 0 5
10
15
20
SAMPLING DISTANCE (Meters)
FIG. 3. Effect of droplet size on survival of aerosolized P. syringae. P. syringae was aerosolized at 22°C and 39% relative humidity as described in the text. Samples were collected and enumerated as described in the text. Large droplet size, Median droplet diameter of 450 p.m. Small droplet size, Median droplet diameter of 130 p.m. Each data point represents the mean + 1 standard error of 12 replicates. Absence of error bars indicates that the amount of error was too small to fit in the scale of this graph.
(b)
120CO
I22°C
,
.-
0
0
*
1~ ~ ~ ~
60
ar,
0.5+
20
15
0
0~~~~
Li
1.0
10
SAMPLING DISTANCE (Meters)
FIG. 2. Comparative survival of bacteria in aerosols. Target organisms were grown in LB broth containing the appropriate antibiotics, washed, and sprayed at a concentration of approximately 107 cells per ml, as described in the text. Aerosol samples were collected at distances of 1, 3, 5, 10, and 15 m and enumerated as described in the text. Each data point represents the mean + 1 standard error of six replicates.
Large Droplet Size 0 Small Droplet Size 0
5
0
20
4.0
1-I
z ,
I0 0
5
10
15
20
SAMPLING DISTANCE (Meters)
FIG. 4. Effect of ambient temperature on survival of aerosolized
P. syringae. P. syringae was aerosolized as described in the text. Samples were collected and enumerated as described in the text. (a) Large droplet size, Median droplet diameter of 450 p.m. (b) Small droplet size, Median droplet diameter of 130 p.m. Each data point represents the mean + 1 standard error of 12 replicates. Absence of error bars indicates that the amount of error was too small to fit in the scale of this graph.
ambient spray temperature of 22°C (Fig. 4a). The relative survival at the lower temperature (12°C) was greatly enhanced when a smaller droplet size was used (Fig. 4b). The experiments described above were done with suspensions of P. syringae which were washed three times in sterile 10 mM phosphate buffer prior to spraying. Unwashed organisms sprayed as large droplets showed no significant decrease in numbers over the entire 15-m sampling distance. However, unwashed organisms sprayed as small droplets showed significantly reduced target/spore ratios at distances of 5, 10, and 15 m (Fig. 5).
DISCUSSION Experiments with four different gram-negative bacteria (Enterobacter cloacae, Erwinia herbicola, K. planticola, and P. syringae) demonstrated that, under similar release conditions (22°C and 39% relative humidity), the different organisms showed different degrees of survival during aerosolization. Erwinia herbicola showed the highest level of survival at ambient temperatures of 22°C (Fig. 2). There were no significant differences in the survival of the other three organisms. Under our test conditions, the detrimental effects of aerosolization on the survival of washed P. syringae were
APPL. ENVIRON. MICROBIOL.
MARTHI ET AL.
3466
3.0 0
Large Droplet Size 0 Small Droplet Size 0
2.5
0 0
0 Q~
2.0 -\
.L
0.
LL-
1.5 0-
0
1.0* O
z
L'-
O
0.5
0'.00
t 5
10
15
2(10
SAMPLING DISTANCE (Meters)
FIG. 5. Survival of unwashed P. syringae during aerosolization. P. syringae was aerosolized at 22°C and 39% RH as described in the text. Droplet sizes are the same as in the legends to Fig. 3 and 4. Each data point represents the mean + 1 standard error of 12 replicates. Absence of error bars indicates that the amount of error was too small to fit in the scale of this graph.
independent of the size of the droplets carrying the bacteria at a higher ambient temperature (22°C; Fig. 4). At an ambient temperature of 22°C and a relative humidity of 39%, the rate of evaporation of droplets is very rapid, even for large droplets (10 to 30 sec; W. Yates, personal communication). Upon evaporation, bacterium-carrying droplets reach a critical final size beyond which there is no further reduction in size. This critical size is dependent on ambient environmental conditions (B. Lighthart, B. T. Shaffer, B. Marthi, and L. Ganio, unpublished data). The absence of any significant differences in the survival of P. syringae between the two droplet sizes at 22°C is probably indicative of the fact that the final particle size in both cases is similar. The rate of evaporation, which is a function of ambient temperature and relative humidity, is reduced at the lower temperature (12°C), since the relative humidity is high (77% under the conditions tested). A relative humidity in the range of 70 to 80% has been shown to have a protective effect on aerosolized bacteria (8, 11). The protective effect seen in Fig. 4 may, therefore, be a combined effect of lower temperature and higher relative humidity. When experiments were conducted with small droplet sizes at an ambient temperature of 12°C, results showed a higher relative target/spore ratio at distances of 10 and 15 m than at distances of 3 and 5 m. It is thought that this may be due to clumping together of B. subtilis spores, used as controls in all experiments, due to the effects of the increased humidity. This may underestimate the number of spores counted and also the relative ratio of target organisms to spores. A high relative humidity may also cause clumping together of vegetative cells and possibly increase survival.
Preparative procedures prior to aerosolization had a profound effect on the survival of P. syringae (Fig. 5). These experiments were done under the same conditions of temperature and relative humidity (22°C and 39%) as the experiments described in the legend to Fig. 2. Unwashed cells sprayed in large droplets did not show any significant decrease in survival over a distance of 15 m (Fig. 5). Washed cells aerosolized under the same conditions showed a significant reduction in survival (Fig. 3). Unwashed cells sprayed as small droplets, however, showed significantly reduced survival. The presence of organic matter from the growth medium (LB broth) increases the vapor pressure of droplets.
This might reduce evaporation and confer protection to the aerosolized bacteria. However, this protective effect may be masked by the more rapid evaporation of smaller droplets, thus reflecting the greater die-off seen in Fig. 5. Due to the short (20-min) collection time in the AGIs and the absence of any nutrient medium (cells are collected in phosphate buffer), the possibility that cells may be multiplying is remote. Additional experiments were done to determine whether the washing procedure caused any reduction in the viable counts of these bacteria. Viable counts of an unwashed suspension of P. syringae were compared with viable counts of the organism washed three times with 10 mM phosphate buffer (pH 7.2). Results showed that there was no significant difference in the viable counts of unwashed and washed cells (data not shown). This indicates that the washing procedure itself did not affect viable counts of these bacteria. Thus, any differences seen in the survival of washed and unwashed bacteria after aerosolization were due to the stress of the aerosolization process itself. In summary, of the four genera used in this study, Erwinia herbicola showed the highest degree of survival during aerosolization over distances of 15 m. There was no significant difference in the survival of Enterobacter cloacae, K. planticola, and P. syringae. The survival of aerosolized P. syringae was found to be dependent on a number of factors, including droplet size, ambient temperature, and relative humidity, and culture preparative procedures. This study is useful for the development of monitoring protocols for detection of released microorganisms, including genetically engineered microorganisms. Such protocols would include the use of media and conditions that would better allow stressed microorganisms to be detected. Results from such studies can also be used to model the downwind dispersal of aerosolized bacteria by predicting the effects of various environmental and preparative conditions on the survival of aerosolized bacteria. ACKNOWLEDGMENTS We are grateful to David Zirkle and Lisa Ganio, NSI Technology Services Corp., for help in the statistical analysis of data presented in this manuscript. LITERATURE CITED 1. Beebe, J. M., and G. B. Pirsch. 1958. Response of air-borne species of Pasteurella to artificial radiation simulating sunlight under different conditions of relative humidity. Appl. Microbiol. 6:127-138. 2. Benbough, J. E. 1967. Death mechanisms in airborne Escherichia coli. J. Gen. Microbiol. 47:325-333. 3. Blakeman, J. P., and N. J. Fokkema. 1982. Potential for biological control of plant diseases on the phylloplane. Annu. Rev. Phytopathol. 20:167-192. 4. Cox, C. S. 1966. The survival of Escherichia coli sprayed into air and into nitrogen from distilled water and from solutions of protecting agents, as a function of relative humidity. J. Gen. Microbiol. 44:15-31. 5. Cox, C. S. 1968. The aerosol survival and cause of death of Escherichia coli K12. J. Gen. Microbiol. 54:169-175. 6. Cox, C. S. 1969. The cause of loss of viability of airborne Escherichia coli K12. J. Gen. Microbiol. 57:77-80. 7. Cox, C. S. 1971. Aerosol survival of Pasteurella tularensis disseminated from the wet and dry states. Appl. Microbiol. 21:482-486. 8. Cox, C. S. 1987. Relative humidity and temperature, p. 172-205. In The aerobiological pathway of microorganisms. John Wiley & Sons, Inc., New York. 9. Cox, C. S., and F. Baldwin. 1967. The toxic effect of oxygen
VOL. 56, 1990
the aerosol survival of Escherichia coli B. J. Gen. Microbiol. 49:115-117. Dorsey, E. L., R. F. Berendt, and E. L. Neff, Jr. 1970. Effect of sodium fluorescein and plating medium on recovery of irradiated Escherichia coli and Serratia marcescens from aerosols. Appl. Microbiol. 20:834-838. Dunklin, E. W., and T. T. Puck. 1948. The lethal effect of relative humidity on air-borne bacteria. J. Exp. Med. 87:87-101. Ehrlich, R., S. Miller, and R. L. Walker. 1970. Relationship between atmospheric temperature and survival of airborne bacteria. Appl. Microbiol. 19:245-249. Hatch, M. T., and R. L. Dimmick. 1966. Physiological response of airborne bacteria to shifts in relative humidity. Bacteriol. Rev. 30:597-603. Hess, G. E. 1965. Effects of oxygen on aerosolized Serratia marcescens. Appl. Microbiol. 13:781-787. Lindow, S. E. 1985. Ecology of Pseudomonas syringae relevant to the field use of Ice-deletion mutants constructed in vitro for plant frost control, p. 23-35. In H. 0. Halvorson, D. Pramer, and M. Rogul (ed.), Engineered organisms in the environment: scientific issues. American Society for Microbiology, Washington, D.C. Lindow, S. E., G. Knudsen, R. J. Seidler, M. V. Walter, V. M.
upon
10.
11.
12. 13. 14.
15.
16.
SURVIVAL OF BACTERIA DURING AEROSOLIZATION
3467
Lambou, P. S. Amy, D. Schmedding, V. Prince, and S. Hern. 1988. Aerial dispersal and epiphytic survival of Pseudomonas syringae during a pretest for the release of genetically engineered strains into the environment. Appl. Environ. Microbiol.
54:1557-1563. 17. Miller, W. S., R. A. Scherff, C. R. Piepoli, and L. S. Idoine. 1961. Physical tracers for bacterial aerosols. Appl. Microbiol. 9:248-251. 18. Poon, C. P. C. 1966. Studies on the instantaneous death of airborne Escherichia coli. Am. J. Epidemiol. 84:1-9. 19. Webb, S. J. 1959. Factors affecting the viability of air-borne bacteria. I. Bacteria aerosolized from distilled water. Can. J. Microbiol. 5:649-669. 20. Webb, S. J. 1960. Factors affecting the viability of air-borne bacteria. II. The effect of chemical additives on the behavior of air-borne cells. Can. J. Microbiol. 6:71-87. 21. Webb, S. J. 1960. Factors affecting the viability of air-borne bacteria. III. The role of bonded water and protein structure in the death of air-borne cells. Can. J. Microbiol. 6:89-105. 22. Webb, S. J. 1967. The influence of oxygen and inositol on the survival of semidried microorganisms. Can. J. Microbiol. 13: 733-742.
