Alberto E. Parma

Virulence genotypes and serotypes of verotoxigenic Escherichia coli isolated from cattle and foods in Argentina

Virulence factors of Verotoxin-producing Escherichia coli(VTEC) strains isolated from hamburgers and ground beef were studied in Argentina by PCR. Their virulence profiles were correlated with those corresponding to strains isolated from calves and adult cattle. Most virulent profiles (VTs+eae+Mp+) were present in E. colifrom healthy and diarrheic calves corresponding to O5:H-, O5:H27, O20:H?, O26:H11, O38:H?, O103:H-, O103:H2, O111:H-, O118:H16, O165:H-serotypes. The presence of the eaegene was significantly more frequent among VTEC strains isolated from calves (20/26; 76%) than from adult cattle (1/39; 2.5%) (p< 0.005). VT2+eae−E. coliwas prevalent in foods and adult cattle at slaughterhouse. The prevalence of the eaegene was similar between VTEC strains isolated from meat (0/21) and adult cattle (1/39; 2.5%) which constitutes the main population processed at slaughterhouses in Argentina. Serotyping showed that VTEC strains were distributed among 31 serotypes, some of which (O20:H19, O91:H21, O113:H21, O116:H21, O117:H7, O171:H2, OX3:H21) were shared between bovine and food strains. These O serogroups have been isolated from cases of haemorrhagic colitis (HC) and haemolytic-uraemic syndrome (HUS) in humans in several continental European countries. This study confirms the role of cattle as a reservoir of many VTEC serotypes other than O157:H7 and represents a base for future diagnostic, prevention and control strategies of EHEC in this country. In addition, this study affirms the advantages of PCR-based screening of E. coliisolates given the finding of so many verotoxin-producing strains.

Serotypes and virulence genes of bovine Shigatoxigenic Escherichia coli (STEC) isolated from a feedlot in Argentina

Grazing-fed cattle were previously demonstrated to be reservoir of non-O157 Shigatoxigenic Escherichia coli (STEC) serotypes in Argentina. The acid-resistance of some STEC strains makes it reasonable to assume the presence in feedlot of particular STEC serotypes. Fifty-nine animals were sampled every 2 weeks during 6 months by rectal swabs. Twenty-seven of 59 animals (45.8%) were shown to be Stx2(+); 3/59 (5.1%) carried Stx1(+) and 7/59 (11.9%) were Stx1(+) Stx2(+). Among 44 STEC isolates, 31 isolates were associated to 10 O serogroups (O2, O15, O25, O103, O145, O146, O157, O171, O174, O175) and 13 were considered non-typable (NT). Six H antigens (H2, H7, H8, H19, H21, H25) were distributed in 21 isolates whereas 23 were non-mobile (H-). Seventeen of 44 strains (38.6%) were eaeA(+) and 14 (31.8%) harbored the 60MDa plasmid. The megaplasmid (Mp) and eaeA gene were simultaneously found in a limited number of serotypes belonging to the enterohaemorrhagic E. coli (EHEC). E. coli O157:H7 strains, isolated from four (6.8%) animals, corresponded to the Stx2(+), eaeA(+), Mp(+) pattern. Three O157:H7 strains belonged to phage type 4 and the other strain was atypical. Many serotypes isolated from grain-fed cattle (O2:H25, O15:H21, O25:H19, O145:H-, O146:H-, O146:H21, O157:H7, O175:H8) also differed from those isolated by us previously from grazing animals. The serotypes O15:H21, O25:H19 and O175:H8 had not been identified at present as belonging to STEC. This work provides new data for the understanding of the ecology of STEC in grain-fed cattle and confirms that cattle are an important reservoir of STEC.

Prevalence of bovine verotoxin-producing Escherichia coli in Argentina.

Faecal swabs obtained from 126 calves and 118 cows in Argentina were investigated for the presence of verotoxin-producing Escherichia coli (VTEC). VTEC strains were recovered from 10 (23%) of 43 calves with diarrhoea, from 24 (29%) of 83 healthy calves, from 40 (44%) of 91 healthy cows waiting at the slaughterhouse, and from 6 (22%) of 27 healthy grazing cattle. PCR showed that 21 (9%) of animals carried VT1+ strains, 49 (20%) VT2+ strains and 10 (4%) VT1+ VT2+ strains. VT1+ strains predominated among calves (16% versus 0.8%; p < 0.001). The presence of eae gene was significantly more frequent among VTEC strains isolated from calves (78%; 46/59) than from cows (2%; 1/65) (p < 0.001). Furthermore, eae gene was more prevalent in VT1+ strains (97%; 32/33) than in VT2+ strains (14%; 10/70) (p < 0.001) and in VT1+ VT2+ strains (24%; 5/21) (p < 0.001). Sorbitol negative high virulent strains serogroups O157 were not detected. This study indicates that cattle are a reservoir of VTEC strains, and that eae gene is associated with VT1+ strains that are predominating among young animals. Fortunately, only adult animals are taken to the slaughterhouse, among which VTEC strains negative for eae gen are predominating.

Lethal leptospiral pulmonary hemorrhage: an emerging disease in Buenos Aires, Argentina.

To the Editor: In the Buenos Aires metropolitan area, 40–100 cases of human leptospirosis are reported annually. Occasional epidemic outbreaks have been characterized by mild leptospiral illness. Severe illness with acute renal failure and extensive cutaneous and visceral hemorrhages (always accompanied by jaundice) has been observed only rarely. A review of our data for 1990–1999 showed that 276 human cases were diagnosed; 43 of these were characterized by pneumonia alone or associated with another syndrome. No severe pulmonary hemorrhage due to leptospirosis was detected in these cases (Table), and the case-fatality rate was <1% (1). Table Clinical findings in human leptospirosis, Hospital F.J. Muniz, Buenos Aires, 1990–1999 Rodents and dogs are considered major reservoirs for this zoonotic illness. Rattus norvegicus (78%) and R. rattus (22%) are the most widely distributed and predominant species. Rodent abundance has been estimated by the Hayne’s Index1 as 0.414–0.465. Prevalence of leptospiral infection as measured by kidney culture of captured rodents ranges from 25% to 40% (1). Antibody prevalence in dogs in Buenos Aires can be as high as 60%. Canine infection is mainly related to the presence of stagnant water and time spent outdoors (2). Statistically, the most important sources of infection are leisure activities (31.4%); certain types of work, including garbage collection, sewer and construction work, and gardening (26.1%); and floods (16.1%) (3). During 2000–2001, a total of 93 cases were reported in this area. An outbreak that included 47 cases took place in March 2001, in Quilmes in the suburban area (Informe de Epidemiologia de Quilmes, Buenos Aires, unpub. data). Four patients died with suspected leptospiral illness; three of these patients had laboratory-confirmed cases. We describe two cases with lethal pulmonary hemorrhage. On July 2000 and March 2001, two women, ages 28 and 34, who lived in urban slum settlements, became ill. A high abundance of rodents inside their houses and in the neighborhood was reported in both cases. After 7–10 days of unspecific febrile illness, a severe pneumonia developed in both women. No jaundice, renal involvement, or thrombocytopenia was observed. When the patients were admitted to the critical-care unit, electrocardiograms were normal for both. For one of the patients, empiric treatment was begun with 4 g of ceftriaxone plus 1 g of erythromycin daily. In the other, 800 mg/day of ciprofloxacin replaced the erythromycin. Endoscopic examination showed no lesions within the bronchial lumen, and abundant hemorrhagic secretions were obtained by aspiration. Both patients were mechanically ventilated and remained stable for the first 48 hours. Between the second and third day of ventilation, they became hypoxemic with acidosis and hypotension. Except for pulmonary hemorrhages, no other sign of bleeding was observed. Both patients died with cardiovascular collapse 10–11 days after onset of illness. The microagglutination test with 10 serovars was positive for leptospirosis, as well as macroagglutination and enzyme-linked immunosorbent assay (ELISA) with leptospiral antigen, for immunoglobulin (Ig) M. Blood, urine, and bronchoalveolar lavage culture were negative for leptospira, as well as for other bacteria. IgM-capture ELISA (Andes serotype) for hantavirus was negative. Pathologic studies performed in one of the patients showed severe hemorrhage inside the pulmonary alveoli, with few interstitial lymphocytes; some septum tissue showed minimal enlargement. Warthin-Starry staining was negative for leptospira. Rodents were captured near one patient’s house, and their kidneys were cultured in Ellinghousen-McCullough-Johnson-Harris medium. Three strains of Leptospira interrogans serovar icterohaemorrhagiae were isolated and characterized; laboratory guinea pigs were injected with the strains and several died 8–10 days later. Tegumentary jaundice was present, as well as abdominal hemorrhage foci. Pulmonary hemorrhages were observed bilaterally. Pericardial hemorrhages are remarkable as a possible cause of cardiopulmonary collapse. Leptospira were recovered from the liver and the kidneys, although brain and lung cultures were negative. Another group of guinea pigs that had also been injected with Leptospira was humanely killed as soon as symptoms appeared. Necropsy showed primary lung injury. Lungs were pale with hemorrhages widely spread over the surface. Lesions were similar to those observed in one of the patients. Neither jaundice nor renal damage was found. Leptospira was isolated from kidneys, lungs, and brain. Jaundice has been reported in severe forms of human disease. Thrombocytopenia has been associated with renal failure and death in human patients. Respiratory involvement in leptospirosis could be classified as a) mild to moderate (20% to 70% of patients), with pulmonary infiltrates commonly associated with jaundice and minimal alteration of renal function; b) severe, with jaundice, nephropathy, and hemorrhages (severe Weils syndrome) (4); and c) fatal, with death occurring as a result of renal failure, myocarditis, or massive pulmonary hemorrhages with cardiovascular collapse. In the past two decades, an increasing number of cases of leptospiral pulmonary hemorrhages have been reported, especially from Southeast Asia (5). In a review of leptospirosis in Brazil, death was associated with renal failure in 76.2% of fatal cases, while 3.5% were related to pulmonary hemorrhages (6). In the epidemic outbreak in Nicaragua in 1995, this form was considered the cause of death in the 40 fatal cases reported (7). The two cases reported here were associated with pulmonary hemorrhage. This clinical form has not been previously reported in the Buenos Aires metropolitan area. Environmental and social factors, the prevalence of infection in reservoirs, and the virulence of the isolated strains must be considered in primary or critical-care units in the diagnosis of new cases, whether or not associated with an outbreak.

Experimental demonstration of an antigenic relationship between Leptospira and equine cornea

Horses inoculated with either equine cornea or killed Leptospira interrogans serovars pomona, tarassovi, icterohaemorrhagiae, wolffi and hardjo, developed corneal opacity and produced antibodies which made it possible to demonstrate partial antigenic identity between equine cornea and four of those serovars employed. These antibodies were isolated by means of immunoadsorptions, purified by ion-exchange chromatography (DEAE-Sephadex A-50) and run by immuno-electrophoresis in agar gel. Both antibodies, anti-equine cornea and anti-leptospira, showed that they corresponded to the IgGb subclass. They bound themselves to equine cornea in vivo and in vitro as was proved by immunofluorescence. This antigenic relationship may be in part responsible for pathogenesis of corneal opacity in leptospirosis of horses.

Role and clinical course of verotoxigenic Escherichia coli infections in childhood acute diarrhoea in Argentina

The aim of this study was to investigate the role and clinical course of verotoxigenic Escherichia coli (VTEC) infections in children with acute diarrhoea from Argentina, the country with the highest worldwide incidence of haemolytic uraemic syndrome (HUS). To accomplish this objective, 437 samples from children up to 6 years old with acute diarrhoea were collected and processed. More than 60 % of the children studied presented watery or mucous diarrhoea without blood, and in 25.2 % of the cases the samples contained blood. In a first screening, a multiplex PCR was performed to detect the presence of the vt(1), vt(2), eae, ehxA and saa virulence genes. The strains were then isolated and analysed to characterize their serotypes, virulence genes, antibiotic susceptibility profiles and verotoxin (VT) production. Forty-four of the 437 samples (10.1 %) were positive for VTEC virulence genes. VTEC-infected patients presented different types of diarrhoea (27.3 % belonged to the non-bloody type). Several serotypes and virulence genotypes were found. Isolates belonged to the serotypes O157 : H7, O145 : H(-), O26 : H11, O121 : H19, O111 : H2 and O118 : H2. HUS developed in 16 (36.4 %) patients positive for VTEC virulence genes. All of the VTEC isolates produced a cytopathic effect on Vero cell monolayers, confirming the ability to express VT. Despite most strains being sensitive to all of the antimicrobials studied, a positive association between clinical progression to HUS and antibiotic therapy was observed for the total number of patients studied, as well as for the VTEC(+) group. In conclusion, the data obtained in this study increase our knowledge of the role and clinical course of VTEC infection in childhood acute diarrhoea beyond bloody diarrhoea, and might be considered for the prevention, diagnosis and management of this disease. It is possible that the optimal approach for VTEC diagnosis could be using multiplex PCR to search for the presence of the vt(1), vt(2), eae and ehxA genes.

A DNA fragment of Leptospira interrogans encodes a protein which shares epitopes with equine cornea

Horses infected with Leptospira interrogans present several clinical disorders, one of them being recurrent uveitis. An antigenic relationship between this bacterium and equine cornea has been described in previous studies. With the aim to make progress on defining the molecular basis and pathogenesis of equine recurrent uveitis, here we describe the cloning of one DNA fragment from a Leptospira interrogans serovar pomona genomic lambda gt11 library. Although there are references of transcription of leptospiral genes in E. coli from their own leptospiral promoters, in this recombinant construction the leptospiral DNA was located under the control of lacZ promoter since no expression could be detected in the absence of IPTG. This clone, isolated by expression screening with polyclonal serum raised against equine corneal proteins, encodes a 90 kDa protein of L. interrogans which crossreacts with equine cornea as proved Western-blotting. Antibodies directed against this leptospiral protein strongly recognised a 66 kDa equine corneal protein, one of those recognised by an anti-equine cornea serum. Our findings suggest that an immune response to 90 kDa protein participates in pathogenesis of equine uveitis.

Tears and aqueous humor from horses inoculated with Leptospira contain antibodies which bind to cornea.

An antigenic relationship between Leptospira interrogans and equine cornea was previously described by us. An enzyme-linked immunosorbent assay was employed in the present work to investigate the existence of anti-leptospira and anti-cornea antibodies in tears, aqueous humor and serum from horses inoculated i.m. with those antigens. Ten days after a booster by the same route, antibodies that bind to microtiter plates, coated with an homogenate of either equine cornea or leptospira, were detected in those fluids and in the sera. At the same time, the corneas of the horses began to develop a diffuse opacity. This finding of anti-leptospira antibodies in equine tears and aqueous humor shows the pathway along which they arrive at the cornea and bind to it.

Toxigenic Escherichia coli isolated from pigs in Argentina

The presence of porcine toxigenic E. coli (ETEC, VTEC) in 28 piggeries (5% of total) of the central and northeast region of Argentina was studied for a better understanding of the epidemiology of porcine strains. Samples were taken by rectal swabs from healthy piglets and from those with diarrhoea, in addition to their dams. Between 5-10 colonies were isolated from each one of 223 animals sampled from 1992 to 1997. By using specific primers each strain was screened by PCR for VT1, VT2all, VT2e, STIa, and LTI toxin genes. Only strains positive for any of the toxins mentioned above were screened for STb. Their O serogroups were determined by agglutination. All of the above enterotoxins and verocytotoxins were found in E. coli isolated from the animals. The STIa gene was detected in E. coli isolated from 27/127 piglets with diarrhoea, in comparison with LTI (4/127 pigs). No toxin gene was amplified from E. coli isolated from either healthy piglets or their dams. When strains isolated from 48 piglets without diarrhoea but showing delayed growth were analysed by PCR, their toxin profile was determined to be VT1 (1/48 piglets), VT2all (5/48), STIa (1/48), LTI (3/48) and VT2e (3/48). Serogroup O64 prevailed among ETEC; O138 prevailed for ETEC/VTEC strains. This is the first extensive study regarding porcine toxigenic E. coli in Argentina and constitutes an important database for the implementation of prevention measures.

Detection of an antigenic protein of Leptospira interroganswhich shares epitopes with the equine cornea and lens

A protein epitope which is involved in an antigenic relationship between equine ocular tissues and Leptospira interrogans was detected in homogenates of the bacterium. The antigenic determinant was harboured on a peptide structure which was shown to be sensitive to the action of denaturing and reducing agents by means of Western blotting. The outer surface of the leptospires appeared to be free of this epitope as was proved by dot-blot and electron microscopic studies.