<italic>Escherichia coli</italic> has been the focus of immense international research driven by its recognition as a major cause of large scale epidemics and tens of thousands of sporadic cases of gastrointestinal illness in animals and man. Therefore, one of the most important areas of current exploration is to understand the virulence attributes of <italic>E. coli</italic> for the development of specific prophylactic and therapeutic measures adopted to combat the diseases. Keeping this in view, <italic>E. coli</italic> isolates recovered from diarrhoea of domestic animals and poultry were subjected to different virulence tests such as congo red (CR) dye binding test, haemagglutination (HA), haemolysin test, enterotoxin and verocytotoxin assays.Congo red (CR) binding property was detected as described by Berkhoff and Vinal (<xref ref-type="bibr" rid="bib002">1986</xref>) with slight modification. Fifty three <italic>E. coli</italic> isolates were streaked on trypticase soya agar which was incorporated with 0.005% CR dye and 0.15% bile salt and incubated at 37°C for 24 h. The plates were further incubated at room temperature for 24 h. The colonies were examined after 18, 24 and 48 h incubation. <italic>E. coli</italic> which produced red colonies within 18 to 48 h of incubation were recorded as CR positive and those which produced grayish white colonies and remained so throughout the incubation period were recorded as CR negative.Haemagglutination (HA) property was detected by a technique described by Veere Gowda <italic>et al</italic>. (<xref ref-type="bibr" rid="bib018">1999</xref>). The ability of test strains to haemagglutinate the erythrocytes of different species was tested in 96 well round bottom microtitre plates. The test cultures were incubated in brain heart infusion broth at 37° C and subcultures of pellicular growth were made every 48 to 72 h upto sixth passage. Cultures were certrifuged at 3700 x g for 20 min and bacterial pellets were resuspended in PBS (1M, pH 7.0) for HA assays. <italic>Escherichia coli</italic> haemagglutination tests were performed using 50 μl of each 1% RBC suspension of chicken, bovine and caprine erythrocytes, <italic>E. coli</italic> suspension (8.0x108 cell/ml) and saline. All 53 <italic>E. coli</italic> were tested for haemagglutination against erythrocytes of above three species. Only HA positive samples were tested for mannose sensitivity by using 50 μl 0.1 M mannose solution instead of saline. Results were read after 30 min at 18° to 20°C and again after overnight at 4°C. The <italic>E. coli</italic> isolates which continued to show HA even after adding mannose to the test were considered as mannose resistant (MRHA) and those which could not retain the property of HA after incorporation of mannose were considered as mannose sensitive (MSHA) strains.Haemolysin production was detected on 5% sheep blood agar medium as per the procedure described by Cruickshank <italic>et al</italic>. (<xref ref-type="bibr" rid="bib004">1975</xref>).For detection of heat labile enterotoxin, cell free filtrates of <italic>E. coli</italic> isolates were prepared as per Gianella (<xref ref-type="bibr" rid="bib011">1976</xref>). Rabbit ligated ileal loop assay was conducted to study the effect of heat labile enterotoxin as per Bergdoll (<xref ref-type="bibr" rid="bib001">1988</xref>).For detection of verotoxin, <italic>E. coli</italic> isolates were grown in brain heart infusion broth for 18 h at 37°C on a mechanical shaker. Broth culture was certrifuged at 12000 x g for 15 min at 4°C. The cell pellet was resuspended in the same supernatant. Then the suspension was ultrasonicated for 2 min at 4°C for carrying out complete disruption of maximum number of cells. The sonicated cell suspension was certrifuged at 3000 x g for 15 min and the supernatant fluid (sonic extract) was harvested. The lysate so obtained was sterilised by filtration using cellulose filter (0.2μ). The filtrate was stored at −20°C until assayed for verotoxin.Sonic extracts of <italic>E. coli</italic> were examined for ability to induce cytopathic effects (CPE) in Vero cells. Assay was conducted using confluent layer of Vero cells. Monolayers were prepared in growth medium by seeding one ml of Vero cells (5.0 x 104 cells/ml) and dispensed into 96 well microtitre plates. Uniform growth of Vero cells in the form of monolayer was obtained after incubation at 37°C for 18 h. Then, the medium from all the wells of a plate was discarded and the wells were washed with Dulbecco's PBS (pH 7.4). The sonic extracts of <italic>E. coli</italic> were serially diluted in maintenance medium and 20 μl of each dilution (1:2, 1:5 and 1:10) was inoculated in Vero cell monolayers and later examined under low and high power of microscope for CPE at 12, 24, 48 and 72 h. Verotoxic <italic>E. coli</italic> (VTEC) and non-verotoxic <italic>E. coli</italic> were kept as controls.In the present study, all the <italic>E. coli</italic> isolates obtained from diarrhoeic cases of domestic animals and poultry were found to bind the CR dye. This finding is in concurrence with the studies of Berkhoff and Vinal (1996) who reported a strong correlation between the expression of CR phenotype and virulence of avian <italic>E. coli</italic> and CR binding is known to be associated with the presence of β-D glucon in the bacterial cell wall. Subsequently, Gjessing and Berkhoff (<xref ref-type="bibr" rid="bib012">1989</xref>) opined that CR dye binding could be used as phenotypic marker to distinguish invasive isolates from non-invasive ones.Thirty one (58.49%) of 53 <italic>E. coli</italic> isolates agglutinated erythrocytes of one or more species among chicken, bovine and caprine and showed different HA patterns. Twenty two (70.97%) of 31 haemagglutination positive isolates agglutinated chicken RBC and 18 (58.06%) of those 31 showed mannose sensitive haemagglutination (MSHA) and 4 (12.90%) mannose resistant haemagglutination (MRHA). Roy <italic>et al</italic>. (<xref ref-type="bibr" rid="bib017">2004</xref>) observed 66.6 and 61.1% <italic>E. coli</italic> strains to be MRHA and MSHA, respectively.The HA assays point to the existence of MR adhesions, perhaps adherence fimbriae on ETEC in which MR adhesions selectively facilitate their attachment to intestinal epithelial cells. Evans <italic>et al</italic>. (<xref ref-type="bibr" rid="bib010">1979</xref>) developed HA typing system for the presumptive identification of different types of fimbriae. It was enunciated that adherence fimbriae were widely found in virulent strains of <italic>E. coli</italic> and were responsible for MRHA. It was generally assumed that <italic>E. coli</italic> of intestinal flora produced common pili and HA activity mediated by such pili was inhibited by mannose (MSHA). In principle, it is relatively simple to differentiate virulent strains with adherence pili from <italic>E. coli</italic> of normal flora on the basis of haemagglutinins.On culturing <italic>E. coli</italic> on sheep blood agar, it was observed that 29 (54.72%) of 53 <italic>E. coli</italic> were haemolytic. This coincides with the observations by Patil <italic>et al.</italic> (<xref ref-type="bibr" rid="bib015">1999</xref>). Hussain <italic>et al</italic>. (<xref ref-type="bibr" rid="bib013">2003</xref>) also reported 7.92% haemorrhagic <italic>E. coli</italic> from diarrhoeic animals.Of 22 <italic>E. coli</italic> isolates tested, 13 (59.09%) were found to be enterotoxigenic. A dilatation index of 0.4 and above was taken as positive for enterotoxigenicity (<xref ref-type="bibr" rid="bib005">De <italic>et al</italic>., 1956</xref>). Kapoor <italic>et al</italic>. (<xref ref-type="bibr" rid="bib014">1998</xref>) and Elfaki (<xref ref-type="bibr" rid="bib007">2000</xref>) observed 5.8 and 30% enterotoxigenic <italic>E. coli</italic>, respectively, among <italic>E. coli</italic> isolated from diarrhoeic calves and lambs. Dubey <italic>et al</italic>. (<xref ref-type="bibr" rid="bib006">2000</xref>) reported 55% <italic>E. coli</italic> from diarrhoeic goats to be enterotoxigenic. The results of present study are in concurrence with the above findings. The heat labile toxin (LT) stimulates adenylcyclase resulting in conversion of ATP to cyclic AMP. Latter induces excretion of Cl− and inhibits the adsorption of Na+, causing great fluid losses. Since, loss of fluid and electrolytes along with congestion and haemorrhage of intestines results in considerable mortality of host animals. Enteroxigenicity (Ent) is a virulence character of utmost importance for animal and to man owing its zoonotic nature.Thirteen (24.53%) of 53 <italic>E. coli</italic> isolates tested produced cytotoxic changes in vero cells. In the toxin treated cells the CPE was noticed by 24 h. The changes included, rounding of cells which appeared slightly enlarged, shrivelled and refractile. By the end of 48 h the cells started detaching from the surface and by 72 h nearly 75% of the cells were detached (<xref ref-type="bibr" rid="bib003">Blanco <italic>et al</italic>. 1993</xref>; <xref ref-type="bibr" rid="bib016">Rathore <italic>et al</italic>. 2003</xref>; <xref ref-type="bibr" rid="bib009">Eriksoon <italic>et al</italic>., 2005</xref>).Emery <italic>et al</italic>. (<xref ref-type="bibr" rid="bib008">1992</xref>) conducted verotoxicity studies. It was possible that these strains might exist merely as transient pathogens alongwith normal intestinal flora, which would multiply rapidly and spread by chance, when bird's defence mechanism is weakened. The elaboration of these toxins during colonization perhaps caused enterotoxicity or damage to intestinal mucosa allowing <italic>E. coli</italic> to penetrate and create subsequent colisepticaemic condition.Comparison of virulence factors of <italic>E. coli</italic> isolates is discussed as follows. All the isolates (53) tested took CR dye and were capsule producing. Out of 53 <italic>E. coli</italic> tested 29 (54.72%), 11 (20.75%) and 13 (24.33%) were haemolytic, MRHA and verotoxic, respectively. Of 22 <italic>E. coli</italic> isolates tested, 13 (59.09%) were found to be enterotoxigenic. Out of 29 haemolytic <italic>E. coli</italic>, four (13.79%) were MRHA and five (17.24%) verotoxic. Of 13 enterotoxic <italic>E. coli</italic>, only one (7.69%) was MRHA, 4 (30.76%) verotoxic and 6 (46.15%) were haemolytic. Of 13 (30.76%) verotoxic strains, 4 (30.76%) were enterotoxigenic, 4 MRHA and 5 (38.76%) haemolytic.Since all the strains were capsulated and CR binding, hence, those were pathogenic. The multiple number of virulence characters, e.g., CR binding, MRHA, haemolytic, enterotoxicity and verotoxicity possessed by any single isolate reflects the pathogenicity in the form of adherence, haemolysis, enteropathogenicity, enterotoxigenicity and enteroinvasiveness.</sec></body><back><ack><title>Acknowledgements

IJCMIID

Indian Journal of Comparative Microbiology, Immunology and Infectious Diseases

0790-9320

Indian Association of Veterinary Microbiologists, Immunologists and Specialists in Infectious Diseases

ijcmiid-27-1-016

SHORT COMMUNICATION

Study of virulence factors of Escherichia coli from domestic animals and poultry

Deshmukh

A.S.

Karpe

A.G.

Department of Veterinary Microbiology, College of Veterinary and Animal Sciences, Parbhani-431 402 (M.S.).

1–62006

2714849

<italic>Escherichia coli</italic> has been the focus of immense international research driven by its recognition as a major cause of large scale epidemics and tens of thousands of sporadic cases of gastrointestinal illness in animals and man. Therefore, one of the most important areas of current exploration is to understand the virulence attributes of <italic>E. coli</italic> for the development of specific prophylactic and therapeutic measures adopted to combat the diseases. Keeping this in view, <italic>E. coli</italic> isolates recovered from diarrhoea of domestic animals and poultry were subjected to different virulence tests such as congo red (CR) dye binding test, haemagglutination (HA), haemolysin test, enterotoxin and verocytotoxin assays.Congo red (CR) binding property was detected as described by Berkhoff and Vinal (<xref ref-type="bibr" rid="bib002">1986</xref>) with slight modification. Fifty three <italic>E. coli</italic> isolates were streaked on trypticase soya agar which was incorporated with 0.005% CR dye and 0.15% bile salt and incubated at 37°C for 24 h. The plates were further incubated at room temperature for 24 h. The colonies were examined after 18, 24 and 48 h incubation. <italic>E. coli</italic> which produced red colonies within 18 to 48 h of incubation were recorded as CR positive and those which produced grayish white colonies and remained so throughout the incubation period were recorded as CR negative.Haemagglutination (HA) property was detected by a technique described by Veere Gowda <italic>et al</italic>. (<xref ref-type="bibr" rid="bib018">1999</xref>). The ability of test strains to haemagglutinate the erythrocytes of different species was tested in 96 well round bottom microtitre plates. The test cultures were incubated in brain heart infusion broth at 37° C and subcultures of pellicular growth were made every 48 to 72 h upto sixth passage. Cultures were certrifuged at 3700 x g for 20 min and bacterial pellets were resuspended in PBS (1M, pH 7.0) for HA assays. <italic>Escherichia coli</italic> haemagglutination tests were performed using 50 μl of each 1% RBC suspension of chicken, bovine and caprine erythrocytes, <italic>E. coli</italic> suspension (8.0x108 cell/ml) and saline. All 53 <italic>E. coli</italic> were tested for haemagglutination against erythrocytes of above three species. Only HA positive samples were tested for mannose sensitivity by using 50 μl 0.1 M mannose solution instead of saline. Results were read after 30 min at 18° to 20°C and again after overnight at 4°C. The <italic>E. coli</italic> isolates which continued to show HA even after adding mannose to the test were considered as mannose resistant (MRHA) and those which could not retain the property of HA after incorporation of mannose were considered as mannose sensitive (MSHA) strains.Haemolysin production was detected on 5% sheep blood agar medium as per the procedure described by Cruickshank <italic>et al</italic>. (<xref ref-type="bibr" rid="bib004">1975</xref>).For detection of heat labile enterotoxin, cell free filtrates of <italic>E. coli</italic> isolates were prepared as per Gianella (<xref ref-type="bibr" rid="bib011">1976</xref>). Rabbit ligated ileal loop assay was conducted to study the effect of heat labile enterotoxin as per Bergdoll (<xref ref-type="bibr" rid="bib001">1988</xref>).For detection of verotoxin, <italic>E. coli</italic> isolates were grown in brain heart infusion broth for 18 h at 37°C on a mechanical shaker. Broth culture was certrifuged at 12000 x g for 15 min at 4°C. The cell pellet was resuspended in the same supernatant. Then the suspension was ultrasonicated for 2 min at 4°C for carrying out complete disruption of maximum number of cells. The sonicated cell suspension was certrifuged at 3000 x g for 15 min and the supernatant fluid (sonic extract) was harvested. The lysate so obtained was sterilised by filtration using cellulose filter (0.2μ). The filtrate was stored at −20°C until assayed for verotoxin.Sonic extracts of <italic>E. coli</italic> were examined for ability to induce cytopathic effects (CPE) in Vero cells. Assay was conducted using confluent layer of Vero cells. Monolayers were prepared in growth medium by seeding one ml of Vero cells (5.0 x 104 cells/ml) and dispensed into 96 well microtitre plates. Uniform growth of Vero cells in the form of monolayer was obtained after incubation at 37°C for 18 h. Then, the medium from all the wells of a plate was discarded and the wells were washed with Dulbecco's PBS (pH 7.4). The sonic extracts of <italic>E. coli</italic> were serially diluted in maintenance medium and 20 μl of each dilution (1:2, 1:5 and 1:10) was inoculated in Vero cell monolayers and later examined under low and high power of microscope for CPE at 12, 24, 48 and 72 h. Verotoxic <italic>E. coli</italic> (VTEC) and non-verotoxic <italic>E. coli</italic> were kept as controls.In the present study, all the <italic>E. coli</italic> isolates obtained from diarrhoeic cases of domestic animals and poultry were found to bind the CR dye. This finding is in concurrence with the studies of Berkhoff and Vinal (1996) who reported a strong correlation between the expression of CR phenotype and virulence of avian <italic>E. coli</italic> and CR binding is known to be associated with the presence of β-D glucon in the bacterial cell wall. Subsequently, Gjessing and Berkhoff (<xref ref-type="bibr" rid="bib012">1989</xref>) opined that CR dye binding could be used as phenotypic marker to distinguish invasive isolates from non-invasive ones.Thirty one (58.49%) of 53 <italic>E. coli</italic> isolates agglutinated erythrocytes of one or more species among chicken, bovine and caprine and showed different HA patterns. Twenty two (70.97%) of 31 haemagglutination positive isolates agglutinated chicken RBC and 18 (58.06%) of those 31 showed mannose sensitive haemagglutination (MSHA) and 4 (12.90%) mannose resistant haemagglutination (MRHA). Roy <italic>et al</italic>. (<xref ref-type="bibr" rid="bib017">2004</xref>) observed 66.6 and 61.1% <italic>E. coli</italic> strains to be MRHA and MSHA, respectively.The HA assays point to the existence of MR adhesions, perhaps adherence fimbriae on ETEC in which MR adhesions selectively facilitate their attachment to intestinal epithelial cells. Evans <italic>et al</italic>. (<xref ref-type="bibr" rid="bib010">1979</xref>) developed HA typing system for the presumptive identification of different types of fimbriae. It was enunciated that adherence fimbriae were widely found in virulent strains of <italic>E. coli</italic> and were responsible for MRHA. It was generally assumed that <italic>E. coli</italic> of intestinal flora produced common pili and HA activity mediated by such pili was inhibited by mannose (MSHA). In principle, it is relatively simple to differentiate virulent strains with adherence pili from <italic>E. coli</italic> of normal flora on the basis of haemagglutinins.On culturing <italic>E. coli</italic> on sheep blood agar, it was observed that 29 (54.72%) of 53 <italic>E. coli</italic> were haemolytic. This coincides with the observations by Patil <italic>et al.</italic> (<xref ref-type="bibr" rid="bib015">1999</xref>). Hussain <italic>et al</italic>. (<xref ref-type="bibr" rid="bib013">2003</xref>) also reported 7.92% haemorrhagic <italic>E. coli</italic> from diarrhoeic animals.Of 22 <italic>E. coli</italic> isolates tested, 13 (59.09%) were found to be enterotoxigenic. A dilatation index of 0.4 and above was taken as positive for enterotoxigenicity (<xref ref-type="bibr" rid="bib005">De <italic>et al</italic>., 1956</xref>). Kapoor <italic>et al</italic>. (<xref ref-type="bibr" rid="bib014">1998</xref>) and Elfaki (<xref ref-type="bibr" rid="bib007">2000</xref>) observed 5.8 and 30% enterotoxigenic <italic>E. coli</italic>, respectively, among <italic>E. coli</italic> isolated from diarrhoeic calves and lambs. Dubey <italic>et al</italic>. (<xref ref-type="bibr" rid="bib006">2000</xref>) reported 55% <italic>E. coli</italic> from diarrhoeic goats to be enterotoxigenic. The results of present study are in concurrence with the above findings. The heat labile toxin (LT) stimulates adenylcyclase resulting in conversion of ATP to cyclic AMP. Latter induces excretion of Cl− and inhibits the adsorption of Na+, causing great fluid losses. Since, loss of fluid and electrolytes along with congestion and haemorrhage of intestines results in considerable mortality of host animals. Enteroxigenicity (Ent) is a virulence character of utmost importance for animal and to man owing its zoonotic nature.Thirteen (24.53%) of 53 <italic>E. coli</italic> isolates tested produced cytotoxic changes in vero cells. In the toxin treated cells the CPE was noticed by 24 h. The changes included, rounding of cells which appeared slightly enlarged, shrivelled and refractile. By the end of 48 h the cells started detaching from the surface and by 72 h nearly 75% of the cells were detached (<xref ref-type="bibr" rid="bib003">Blanco <italic>et al</italic>. 1993</xref>; <xref ref-type="bibr" rid="bib016">Rathore <italic>et al</italic>. 2003</xref>; <xref ref-type="bibr" rid="bib009">Eriksoon <italic>et al</italic>., 2005</xref>).Emery <italic>et al</italic>. (<xref ref-type="bibr" rid="bib008">1992</xref>) conducted verotoxicity studies. It was possible that these strains might exist merely as transient pathogens alongwith normal intestinal flora, which would multiply rapidly and spread by chance, when bird's defence mechanism is weakened. The elaboration of these toxins during colonization perhaps caused enterotoxicity or damage to intestinal mucosa allowing <italic>E. coli</italic> to penetrate and create subsequent colisepticaemic condition.Comparison of virulence factors of <italic>E. coli</italic> isolates is discussed as follows. All the isolates (53) tested took CR dye and were capsule producing. Out of 53 <italic>E. coli</italic> tested 29 (54.72%), 11 (20.75%) and 13 (24.33%) were haemolytic, MRHA and verotoxic, respectively. Of 22 <italic>E. coli</italic> isolates tested, 13 (59.09%) were found to be enterotoxigenic. Out of 29 haemolytic <italic>E. coli</italic>, four (13.79%) were MRHA and five (17.24%) verotoxic. Of 13 enterotoxic <italic>E. coli</italic>, only one (7.69%) was MRHA, 4 (30.76%) verotoxic and 6 (46.15%) were haemolytic. Of 13 (30.76%) verotoxic strains, 4 (30.76%) were enterotoxigenic, 4 MRHA and 5 (38.76%) haemolytic.Since all the strains were capsulated and CR binding, hence, those were pathogenic. The multiple number of virulence characters, e.g., CR binding, MRHA, haemolytic, enterotoxicity and verotoxicity possessed by any single isolate reflects the pathogenicity in the form of adherence, haemolysis, enteropathogenicity, enterotoxigenicity and enteroinvasiveness.</sec></body><back><ack><title>Acknowledgements

The authors are grateful to Associate Dean, College of Veterinary and Animal Sciences, Parbhani, Maharastra for providing facilities.

ReferencesBergdoll

M.S.

(1988) Ileal loop fluid accumulation test for diarrhoeal toxins. In: Methods in Enzymology. Vol. 165. Ed. Harshman

, Acad. Press, London.Berkhoff

N.A.

Vinal

A.C.

(1986) Avain Dis., 30: 117.Blanco

Blanco

J.E.

Blanco

(1993) Microbiologia, 9: 72.Cruickshank

Duguid

J.P.

Marmion

B.P.

Swain

R.H.A.

(1975) Medical Microbiology. 12^th edn., Vol. II, Churchill Livingstone, Edinburgh, London and New York, pp. 113.De

S.N.

Bhatacharya

Sarkar

J.K.

(1956) J. Path. Bact., 71: 201.Dubey

Sharda

Chhabra

Sharma

Reddy

A.G.

Sharma

(2000) Indian J. Comp. Microbiol. Immunol. Infect. Dis., 21: 68.Elfaki

M.G.

(2000) Indian Vet. J., 77: 468.Emery

D.A.

Nagraj

K.V.

Shaw

D.P.

Newman

J.A.

White

D.G.

(1992) Avian Dis., 36: 504.Eriksson

Aspan

Gunnarsson

Vagsholm

(2005) Epidemiol. Infect., 133: 349.Evans Dolers

J.D.

Jr.Evans

Herbert

L.D.

(1979) Infect. Immun., 23: 336.Gianella

R.A.

(1976) Infect. Immun., 14: 95.Gjessing

K.M.

Berkhoff

M.A.

(1989) Avian Dis., 33: 473.Hussain Saikia

G.K.

Rahman

(2003) Indian J. Comp. Microbiol. Immunol. Infect. Dis., 24: 63.Kapoor

K.N.

Kulshrestha

S.B.C.

(1998) Indian J. Comp. Microbiol. Immunol. Infect. Dis., 19: 72.Patil

R.M.

Karpe

A.G.

Gujar

M.B.

(1999) Indian J. Comp. Microbiol. Immunol. Infect. Dis., 20: 29.Rathore

R.S.

Bachil

V.H.

Agarwal

R.K.

Kapoor

K.N.

(2003) Indian J. Comp. Microbiol. Immunol. Infect. Dis., 24: 201.Roy

Edwin

P.G.

Purushothaman

(2004) Indian Vet. J., 81: 1317.Veere Gowda

B.M.

Krishna Murthy

G.V.

Raghwan

(1999) Indian J. Comp. Microbiol. Immunol. Infect. Dis., 20: 121.