THE JOURNAL OF INFECTIOUS DISEASES. VOL. 137, NO.5. MAY 1978 © 1978 by the University of Chicago. 0022-1899178/3705-0006$00.81
Some Characteristics of Escherichia coli Strains Isolated from Extraintestinal Infections of Humans From the Departments of Surgery and of Microbiology and Immunology, University of Washington School of Medicine, Seattle, Washington
Barbara H. Minshew, Julius Jorgensen,* Martha Swanstrum, t Grada A. Grootes-Reuvecamp.j and Stanley Falkow
Because Escherichia coli is of such overwhelming significance in urinary tract infections and is often implicated in many other extraintestinal infections, there has been considerable 'interest in the nature of E. coli strains that cause infection. We have surveyed a series of E. coli strains isolated from extraintestinal infections and, for comparison, strains randomly picked from stool cultures of normal individuals and several E. coli KI2 laboratory strains for some virulence-associated characteristics. These characteristics included production of hemolysin and colicin V, hemagglutination of human erythrocytes (HAil)' and the virulence of these E. coli strains for 13day-old chick embryos that were inoculated alThis research was supported by grant no. AI 10885 from the National Institute of Allergy and Infectious Diseases. Please address requests for reprints to Dr. Barbara H. Minshew, Department of Surgery RF-25, University of Washington School of Medicine, Seattle, Washington 98195. '*' Present address: Department of Periodontics, Center for Research in Oral Biology, University of Washington, Seattle, Washington 98195. t Present address: Department of Microbiology, Yakima Valley Memorial Hospital, 2811 Tieton Drive, Yakima, Washington 98206. t Permanent address: Laboratory of Medical Microbiology, University of Groningen, Oostersingel 59, Groningen, The Netherlands.
648
lantoically. Because extrachromosomal elements (plasmids) may determine virulence-associated characteristics, we used agarose gel electrophoresis to analyze the plasmid component of a series of phenotypically characterized E. coli strains isolated from urinary tract infections and normal enteric flora. Materials and Methods
Bacteria. Cultures of E. coli isolated from blood, urine, miscellaneous extraintestinal sources, and normal persons were obtained from the culture collection of this laboratory and have been described [1]. An additional 31 strains isolated from patients with urinary tract infections (UTI) were provided by Dr. George W. Counts (University of Washington, Seattle, Wash.). The characteristics of the E. coli KI2 strains are listed in table 3. Hemolysin production. E. coli strains were tested for f3-hemolysis after growth at 37 C on blood agar base (Baltimore Biological Laboratories [BBL], Cockeysville, Md.) supplemented with 5% sheep blood. Colicin production. Colicin production was detected by an agar overlay method [2]. E. coli strains were inoculated onto trypticase soy agar (TSA) (BBL) with sterile toothpicks. After
Downloaded from http://jid.oxfordjournals.org/ at New York University on May 24, 2015
Escherichia coli strains isolated from extraintestinal infections of humans possess a constellation of phenotypes not usually found in random fecal isolates, enteropathogenic strains, or the laboratory strain E. coli K12. The phenotypes found more commonly inextraintestinal strains include hemolysin production, the biosynthesis of colicin V, and the hemagglutination of human erythrocytes in the presence of D-mannose (HAJJ Hemolysin is assumed to be a cytotoxic factor, colicin V is assumed to interfere with host defense mechanisms, and HAil is assumed to playa role in specific tissue adherence. In addition, ~50% of E. coli strains from extraintestinal infections kill allantoically inoculated 13-day-old chick embryos. Some (20%) of the fecal E. coli also kill chick embryos, but E. coli Kl2 is innocuous in this virulence model. The plasmids for hemolysin and colicin V production have been transmitted to E. coli K12 derivatives but are not sufficient to convert laboratory strains to a form virulent for the chick embryo.
649
Characteristics of E. coli
and the plasmid component was screened by the agarose gel electrophoresis technique of Meyers et al. [4]. Results and Discussion
Properties contributing to pathogenicity. Although E. coli is ordinarily part of the normal enteric flora of humans and animals, it is also the most common cause of UTI and the most frequently isolated facultative gram-negative bacillus in many other extraintestinal infections. Two views have been held regarding the nature of E. coli isolated from extraintestinal infections. One view is that the E. coli causing infection is the most prevalent type in the feces and is simply "in the right place at the right time." The alternative view attributes "special properties" to E. coli that cause infection; these properties permit the organisms to initiate and sustain an extraintestinal infection [5]. Although there are insufficient data to permit a clear choice between the two alternatives, it appears that the latter hypothesis is most often correct for certain UTI [6], and perhaps for other extraintestinal infections as well. Several properties of E. coli are more frequently associated with strains isolated from extraintestinal infection than with random isolates of normal enteric flora. Some of these properties are production of hemolysin [7, 8], biosynthesis of colicin V [9], and adherence to specific target cells [10]. Hemolysin production. The results of several investigations indicate that hemolysin production by E. coli may be an important factor in the pathogenesis of infection. Among E. coli isolated from patients with UTI, a substantial number of strains may be hemolytic. Cooke and Ewins [7] reported an incidence of 56%, but Vahlne [8] found that only 26% of the strains in his study were hemolytic. Hemolytic strains of E. coli are more frequently isolated from patients with peritonitis or appendicitis than from normal enteric flora [8], in which the incidence of hemolytic strains has been estimated to range from 8% [11] to 18% [12]. In our survey of E. coli isolated from extraintestinal sites, the overall incidence of hemolytic strains was 46%; it varied from 35% for blood, to 49% for urine, to 59% for strains isolated from miscellaneous sites (table
Downloaded from http://jid.oxfordjournals.org/ at New York University on May 24, 2015
overnight growth at 35 C, the cultures were treated with chloroform vapor for at least 30 min and then were allowed to air dry for -====30 min by placing the covers of the petri dishes ajar. The cultures were next overlaid with 0.4% agar (Difco, Detroit, Mich.) in nutrient broth (Di£Co) with 0.5% NaCl containing approximately 106 viable cella/rnl of a suspension of E. coli strain 711 or 711CoIV+ (carrying the ColV plasmid, which codes for production of colicin V). Test strains were considered producers of colicin V if they inhibited indicator strain E. coli 711 but not strain 711CoIV+. Hemagglutination of erythrocytes. Hemagglutination of human type 0, Rh-positive erythrocytes was determined by a slide test as described by Powell and Finkelstein [3]. Human blood was collected in 3.8% sodium citrate, the erythrocytes were washed three times with 0.85% N aCl, and the cells were suspended at a 3% (vol/vol) concentration. All reagents were kept on ice during the test. Bacteria from overnight growth on TSA were suspended in a drop of either 0.85% N aCl or 0.85% N aCl containing 0.570 n-mannose on a chilled glass microscope slide. A drop of 3% (volyvol) human erythrocytes in 0.85% N aCl was added to the bacterial suspension, and the slide was rocked gently for observation of macroscopic agglutination. Experimental infections in chick embryos. As a test of virulence, E. coli strains were inoculated allantoically into 13-day-old chick embryos [3]. Embryonated eggs of White Leghorn chickens (Lab Associates, Kirkland, Wash.) were kept at 37 C in a humidified incubator for 24 hr before inoculation. A O.I-ml volume of 0.85% NaCl (for controls) or of E. coli (1-5 X 104 viable cells) suspended in 0.85% N aCl was injected into the allantoic cavity. The bacterial inoculum was prepared by harvest of overnight growth from TSA into 0.85% N aCl. The bacterial suspension was adjusted to an optical density of 0.35 at 550 nm in a Coleman Jr. spectrophotometer and then diluted I: 1,000. Death of the embryos was determined by candling. Strains were considered virulent when ~60% of the embryos died within 72 hr and avirulent when mortality was ~50%. Agarose gel electrophoresis of plasmid components. Lysates of E. coli strains were prepared,
Minshew et al.
650
Table 1. Incidence of hemolysin and colicin V production by isolates of Escherichia coli from various sources [I] .
Source of strains Blood Urine Miscellaneous * Normal feces
Table 2. Virulence for I3-day-old chick embryos and hemagglutinating properties of Escherichia coli isolated from various sources [I] . Percentage of strains positive for
Percentage of strains producing
No. tested
Hemolysin
51 59 32 20
35 49 56 5
Colicin V 12 7 16 0
*Miscellaneous sources included sputum, wounds, secretions, drainage, abscesses, and other extraintestinal sources.
Extraintestinal infection Intestines of normal individuals
No. tested
Virulence *
Hernagglutination'l
22
59
59
20
20
15
*A strain was considered virulent if 10 4 viable cells, inoculated into the allantoic cavity, killed :;;;,60% of 13-day old chick embryos. There was a significant difference in virulence for the chick embryo with strains from extraintestinal sources vs. E. coli isolated from normal feces (unpaired test of proportions, P
Some Characteristics of Escherichia coli Strains Isolated from Extraintestinal Infections of Humans From the Departments of Surgery and of Microbiology and Immunology, University of Washington School of Medicine, Seattle, Washington
Barbara H. Minshew, Julius Jorgensen,* Martha Swanstrum, t Grada A. Grootes-Reuvecamp.j and Stanley Falkow
Because Escherichia coli is of such overwhelming significance in urinary tract infections and is often implicated in many other extraintestinal infections, there has been considerable 'interest in the nature of E. coli strains that cause infection. We have surveyed a series of E. coli strains isolated from extraintestinal infections and, for comparison, strains randomly picked from stool cultures of normal individuals and several E. coli KI2 laboratory strains for some virulence-associated characteristics. These characteristics included production of hemolysin and colicin V, hemagglutination of human erythrocytes (HAil)' and the virulence of these E. coli strains for 13day-old chick embryos that were inoculated alThis research was supported by grant no. AI 10885 from the National Institute of Allergy and Infectious Diseases. Please address requests for reprints to Dr. Barbara H. Minshew, Department of Surgery RF-25, University of Washington School of Medicine, Seattle, Washington 98195. '*' Present address: Department of Periodontics, Center for Research in Oral Biology, University of Washington, Seattle, Washington 98195. t Present address: Department of Microbiology, Yakima Valley Memorial Hospital, 2811 Tieton Drive, Yakima, Washington 98206. t Permanent address: Laboratory of Medical Microbiology, University of Groningen, Oostersingel 59, Groningen, The Netherlands.
648
lantoically. Because extrachromosomal elements (plasmids) may determine virulence-associated characteristics, we used agarose gel electrophoresis to analyze the plasmid component of a series of phenotypically characterized E. coli strains isolated from urinary tract infections and normal enteric flora. Materials and Methods
Bacteria. Cultures of E. coli isolated from blood, urine, miscellaneous extraintestinal sources, and normal persons were obtained from the culture collection of this laboratory and have been described [1]. An additional 31 strains isolated from patients with urinary tract infections (UTI) were provided by Dr. George W. Counts (University of Washington, Seattle, Wash.). The characteristics of the E. coli KI2 strains are listed in table 3. Hemolysin production. E. coli strains were tested for f3-hemolysis after growth at 37 C on blood agar base (Baltimore Biological Laboratories [BBL], Cockeysville, Md.) supplemented with 5% sheep blood. Colicin production. Colicin production was detected by an agar overlay method [2]. E. coli strains were inoculated onto trypticase soy agar (TSA) (BBL) with sterile toothpicks. After
Downloaded from http://jid.oxfordjournals.org/ at New York University on May 24, 2015
Escherichia coli strains isolated from extraintestinal infections of humans possess a constellation of phenotypes not usually found in random fecal isolates, enteropathogenic strains, or the laboratory strain E. coli K12. The phenotypes found more commonly inextraintestinal strains include hemolysin production, the biosynthesis of colicin V, and the hemagglutination of human erythrocytes in the presence of D-mannose (HAJJ Hemolysin is assumed to be a cytotoxic factor, colicin V is assumed to interfere with host defense mechanisms, and HAil is assumed to playa role in specific tissue adherence. In addition, ~50% of E. coli strains from extraintestinal infections kill allantoically inoculated 13-day-old chick embryos. Some (20%) of the fecal E. coli also kill chick embryos, but E. coli Kl2 is innocuous in this virulence model. The plasmids for hemolysin and colicin V production have been transmitted to E. coli K12 derivatives but are not sufficient to convert laboratory strains to a form virulent for the chick embryo.
649
Characteristics of E. coli
and the plasmid component was screened by the agarose gel electrophoresis technique of Meyers et al. [4]. Results and Discussion
Properties contributing to pathogenicity. Although E. coli is ordinarily part of the normal enteric flora of humans and animals, it is also the most common cause of UTI and the most frequently isolated facultative gram-negative bacillus in many other extraintestinal infections. Two views have been held regarding the nature of E. coli isolated from extraintestinal infections. One view is that the E. coli causing infection is the most prevalent type in the feces and is simply "in the right place at the right time." The alternative view attributes "special properties" to E. coli that cause infection; these properties permit the organisms to initiate and sustain an extraintestinal infection [5]. Although there are insufficient data to permit a clear choice between the two alternatives, it appears that the latter hypothesis is most often correct for certain UTI [6], and perhaps for other extraintestinal infections as well. Several properties of E. coli are more frequently associated with strains isolated from extraintestinal infection than with random isolates of normal enteric flora. Some of these properties are production of hemolysin [7, 8], biosynthesis of colicin V [9], and adherence to specific target cells [10]. Hemolysin production. The results of several investigations indicate that hemolysin production by E. coli may be an important factor in the pathogenesis of infection. Among E. coli isolated from patients with UTI, a substantial number of strains may be hemolytic. Cooke and Ewins [7] reported an incidence of 56%, but Vahlne [8] found that only 26% of the strains in his study were hemolytic. Hemolytic strains of E. coli are more frequently isolated from patients with peritonitis or appendicitis than from normal enteric flora [8], in which the incidence of hemolytic strains has been estimated to range from 8% [11] to 18% [12]. In our survey of E. coli isolated from extraintestinal sites, the overall incidence of hemolytic strains was 46%; it varied from 35% for blood, to 49% for urine, to 59% for strains isolated from miscellaneous sites (table
Downloaded from http://jid.oxfordjournals.org/ at New York University on May 24, 2015
overnight growth at 35 C, the cultures were treated with chloroform vapor for at least 30 min and then were allowed to air dry for -====30 min by placing the covers of the petri dishes ajar. The cultures were next overlaid with 0.4% agar (Difco, Detroit, Mich.) in nutrient broth (Di£Co) with 0.5% NaCl containing approximately 106 viable cella/rnl of a suspension of E. coli strain 711 or 711CoIV+ (carrying the ColV plasmid, which codes for production of colicin V). Test strains were considered producers of colicin V if they inhibited indicator strain E. coli 711 but not strain 711CoIV+. Hemagglutination of erythrocytes. Hemagglutination of human type 0, Rh-positive erythrocytes was determined by a slide test as described by Powell and Finkelstein [3]. Human blood was collected in 3.8% sodium citrate, the erythrocytes were washed three times with 0.85% N aCl, and the cells were suspended at a 3% (vol/vol) concentration. All reagents were kept on ice during the test. Bacteria from overnight growth on TSA were suspended in a drop of either 0.85% N aCl or 0.85% N aCl containing 0.570 n-mannose on a chilled glass microscope slide. A drop of 3% (volyvol) human erythrocytes in 0.85% N aCl was added to the bacterial suspension, and the slide was rocked gently for observation of macroscopic agglutination. Experimental infections in chick embryos. As a test of virulence, E. coli strains were inoculated allantoically into 13-day-old chick embryos [3]. Embryonated eggs of White Leghorn chickens (Lab Associates, Kirkland, Wash.) were kept at 37 C in a humidified incubator for 24 hr before inoculation. A O.I-ml volume of 0.85% NaCl (for controls) or of E. coli (1-5 X 104 viable cells) suspended in 0.85% N aCl was injected into the allantoic cavity. The bacterial inoculum was prepared by harvest of overnight growth from TSA into 0.85% N aCl. The bacterial suspension was adjusted to an optical density of 0.35 at 550 nm in a Coleman Jr. spectrophotometer and then diluted I: 1,000. Death of the embryos was determined by candling. Strains were considered virulent when ~60% of the embryos died within 72 hr and avirulent when mortality was ~50%. Agarose gel electrophoresis of plasmid components. Lysates of E. coli strains were prepared,
Minshew et al.
650
Table 1. Incidence of hemolysin and colicin V production by isolates of Escherichia coli from various sources [I] .
Source of strains Blood Urine Miscellaneous * Normal feces
Table 2. Virulence for I3-day-old chick embryos and hemagglutinating properties of Escherichia coli isolated from various sources [I] . Percentage of strains positive for
Percentage of strains producing
No. tested
Hemolysin
51 59 32 20
35 49 56 5
Colicin V 12 7 16 0
*Miscellaneous sources included sputum, wounds, secretions, drainage, abscesses, and other extraintestinal sources.
Extraintestinal infection Intestines of normal individuals
No. tested
Virulence *
Hernagglutination'l
22
59
59
20
20
15
*A strain was considered virulent if 10 4 viable cells, inoculated into the allantoic cavity, killed :;;;,60% of 13-day old chick embryos. There was a significant difference in virulence for the chick embryo with strains from extraintestinal sources vs. E. coli isolated from normal feces (unpaired test of proportions, P
