Unit: Cholera and Vibrios
Director: FOURNIER Jean-Michel
The scientific activities of the Unité du Choléra et des Vibrions focus on both cholera and non-cholera vibrios. Our contribution to the fight against cholera, which remains a major international health concern, includes: (1) undertaking surveys and molecular epidemiology studies of cholera; (2) developing a rapid diagnostic test for cholera; (3) research into new, chemically defined conjugate vaccines. We are also actively involved in the study of other (non-cholera) vibrios isolated from humans or seafood. The French Ministry of Health has designated our Research Unit the "National Reference Center for Vibrios and Cholera" because of our expertise in these particular fields and we are in the process of being recognized as a WHO Collaborating Center for Cholera and Other Vibrio Infections.
1. Survey and molecular epidemiology of cholera (Marie-Laure QUILICI)
Cholera remains a major public and individual health problem. According to the WHO, the number of people susceptible to cholera due to socioeconomic problems has increased dramatically worldwide, creating favorable conditions for a global cholera crisis.
Cases of cholera due to Vibrio cholerae O1 or O139 have been reported from all continents. France, like most developed countries, does not suffer directly from cholera epidemics. However, imported cases of cholera are regularly identified in travelers (2 imported cases of cholera in 2005 in France). We therefore collaborate closely with biologists from other countries in which cholera outbreaks occur, with members of the International Network of Pasteur Institutes and with humanitarian non-governmental organizations.
We use molecular typing methods to study V. cholerae O1 strains from diverse geographical areas. An analysis of more than 500 cholera isolates collected over a 35-year period since the arrival of cholera in Africa in 1970 has shown a high degree of genetic similarity between African V. cholerae O1 strains, consistent with the seventh pandemic strains of this continent being clonal. We have also studied more than 150 clinical V. cholerae O1 strains isolated in Mexico from 1991 to 2000. We constructed a database, using the results of these studies, and we use this database actively to follow the evolution of cholera vibrio populations and of new cholera vibrio strains that might be imported into France.
2. Rapid diagnostic test for cholera (Alain BOUTONNIER, Jean-Michel FOURNIER)
Our efforts to develop a conjugate vaccine targeting V. cholerae O1 and O139 has led to the development of monoclonal antibodies specific for O1 or O139 lipopolysaccharide. Together with the Laboratoire d'Ingénierie des Anticorps at the Institut Pasteur, Paris, and the Institut Pasteur of Madagascar, we have used the specificity of these monoclonal antibodies to develop a rapid diagnostic test for V. cholerae O1 and O139. This diagnostic test is based on a one-step immunochromatographic technique and has been evaluated in Bangladesh, Madagascar, and Mozambique. The sensitivity and specificity of our test were evaluated and compared with those of two commercially available rapid diagnostic tests, at a diarrhea treatment center in Dhaka, Bangladesh. Our test was the most sensitive, regardless of the level of skill of the technician. It is the most appropriate rapid diagnostic test for cholera in remote locations or refugee camp settings. This test is now marketed by an Indian company.
3. Research program on new chemically defined conjugate vaccines against cholera (Alain BOUTONNIER, Cyrille GRANDJEAN, Bruno DASSY, Jean-Michel FOURNIER)
Cholera is an epidemic diarrheal disease caused by two serogroups of a bacterium, Vibrio cholerae. We are currently in the seventh pandemic, which began in India in 1961, invaded Africa in 1970 and spread to Latin America in 1991. The seventh pandemic is caused by V. cholerae serogroup O1, biotype El Tor, serotypes Ogawa or Inaba. A new V. cholerae serogroup, O139, appeared in India in 1992. There are definite indications that the incidence of V. cholerae O139 is rising in India and Bangladesh. Thus, this new cholera strain could well become a global threat. Hence, there is an urgent need for cholera vaccines conferring reliable, long-term protection against V. cholerae O1 and V. cholerae O139 in all age groups, including children under five years of age.
We began by determining which antibodies protect against cholera. We showed that immunoglobulin G (IgG) monoclonal antibodies directed against the polysaccharide moiety of the lipopolysaccharide of V. cholerae O1, expressing the serogroup or serotype determinants, were immunoprotective in an experimental model of cholera in neonatal mice. This encouraged us to start developing conjugate vaccines composed of the polysaccharides of V. cholerae O1 and V. cholerae O139 covalently linked to a carrier protein, which should induce a long-lasting, thymus-dependent immune response. We have already prepared a conjugate containing the polysaccharide of V. cholerae O139 and have shown that this vaccine induces a protective thymus-dependent response in mice.
This year, we have encountered difficulties in the production of a V. cholerae O139 conjugate meeting Good Manufacturing Practice (GMP) requirements. These difficulties were due to a lack of reproducibility of the process of lipopolysaccharide detoxification. We have continued our efforts to develop novel glycoconjugates in which the immunogenicity of the antigenic determinant common to both the Ogawa and Inaba serotypes of V. cholerae O1 is preserved. In collaboration with Laurence A. Mulard, Unité de Chimie Organique, we have developed innovative methods of conjugation based on the site-specific binding of the polysaccharide antigen to a protein carrier. During this investigation, we designated new phosphino-functionalized cross-linkers suitable for use in protein carrier derivatization. The protein carriers were then conjugated, via an amide bond, to azido-functionalized polysaccharide, by traceless Staudinger ligation. Although the conjugates were antigenic, none elicited a protective thymus-dependent response in mice.
We have also continued our investigations of the structure of the antigenic determinant common to the Ogawa and Inaba serotypes of V. cholerae O1, using immunochemical and physicochemical methods, in collaboration with the Unité de Biochimie Structurale and the Plate-forme de Cristallogenèse et diffraction des rayons X. The Fab fragment of a monoclonal antibody specific for both the Ogawa and Inaba serotypes was produced by papain digestion and separated by ion-exchange chromatography and gel filtration. Crystals were obtained by the hanging-drop vapor diffusion method. The crystal structure of the Fab fragment was determined at a resolution of 2.8 Å, by molecular replacement. A putative monosaccharide-binding cavity was identified in the recombination site, suggesting that the antibody may specifically bind a terminal sugar from the LPS. Molecular mechanics calculations and cocrystallization trials with various Fab-carbohydrate complexes are currently underway, to identify the actual determinant. The determination of this structure may facilitate the rational design of a V. cholerae O1 conjugate.
4. Molecular characterization of non-cholera vibrios (Annick ROBERT-PILLOT, Marie-Laure QUILICI)
Vibrios are gram-negative bacteria and form part of the normal bacterial flora of aquatic environments. Vibrios can be subdivided into "cholera vibrios", which include isolates belonging to the serogroups O1 and O139 of Vibrio cholerae, and "non-cholera vibrios", which include (i) isolates belonging to serogroups other than O1 and O139 of V. cholerae known as V. cholerae non-O1/non-O139 and (ii) isolates belonging to other Vibrio species, including two species frequently isolated from human clinical samples V. parahaemolyticus and V. vulnificus and four other species less frequently isolated V. alginolyticus, V. fluvialis, V. hollisae and V. mimicus.
Environmental and physiological modifications in certain coastal regions can provide vibrios with ideal conditions for proliferation. The development of international trade and increases in the consumption of raw or lightly cooked seafood, coupled with greater numbers of susceptible people, is causing concern that the incidence of infections in Europe may increase. Therefore, human vibrio infections are likely to spread and microbiological surveillance is needed to control public health risks.
The classic biochemical methods used for bacterial identification are not suitable for studying vibrio strains isolated from the environment. Increasing public health concerns emphasize the importance of developing molecular methods for detection, identification and characterization of the pathogenic factors in vibrio strains isolated from humans, seawater or seafood. We have therefore been developing and using a real-time PCR method to detect and quantify directly, in artificially contaminated environmental samples, two medically important Vibrio species, V. cholerae and V. parahaemolyticus. These assays use species-specific oligonucleotide primers targeting sequences of the 16S-23S rRNA intergenic spacer regions for V. cholerae and the R72H sequence for V. parahaemolyticus, with detection by the SYBR green method. We optimized MgCl2 concentration and developed external DNA controls for validation of the test results. The best results were obtained with samples treated with Instagene matrix (Bio-Rad). The sensitivity was about 50 CFU in the reaction mixture when tested on fresh water seeded with V. cholerae or on oyster mantle fluid seeded with V. parahaemolyticus. The real-time SYBR Green PCR assay was also optimized to detect the cholera toxin genes of V. cholerae, and the thermostable direct hemolysin (tdh) and thermostable related hemolysin (trh) genes, markers of V. parahaemolyticus pathogenicity. In contrast to the results obtained with cholera toxin, we could not confirm the specificity of the various primers described in previous studies for tdh gene detection. Indeed, tdh amplification was also observed with strains of V. cholerae and V. hollisae. The real-time PCR assay described here for evaluating populations of V. cholerae and V. parahaemolyticus in seeded environmental samples was highly sensitive and less time-consuming than conventional methods (MPN) or other PCR protocols. Further experiments should improve the detection limits for the quantification of V. cholerae and V. parahaemolyticus. We will also study tdh-specific primers for the direct detection of pathogenic V. parahaemolyticus in environmental samples.
5. Activity of the National Reference Center for Vibrios and Cholera (Marie-Laure QUILICI, Annick ROBERT-PILLOT, Alain GUENOLE, Laure LEMEE, Jean-Michel FOURNIER)
The NRCVC is responsible for identification of the cholera vibrio strains causing clinical cases of cholera in France, where, as in many other countries, cholera is a notifiable disease. Only a few cholera cases (<5) occur each year in France, and most are imported. Two cholera cases, imported from Pakistan, were identified in 2005. Moreover, as cholera epidemics do not respect national borders, we actively collaborate with biologists from other countries with cholera outbreaks and with humanitarian non-governmental organizations. As part of this collaboration, we have studied 62 V. cholerae O1 strains from African countries. These collaborations facilitate the study of new cholera vibrio strains imported into France.
The NRCVC is also responsible for identifying non-cholera vibrio strains from clinical laboratories. Human non-cholera vibrio infections may result in gastroenteritis, skin and soft-tissue infections, septicemia, and other extraintestinal infections, such as ear infections. Patients with underlying immunosuppressive diseases are at high risk of the rapid spread of infections due to these microorganisms. Most infections are associated with contact with seawater or the consumption of seafood, and their occurrence is correlated with the warmer months of the year.
In 2005, 10 cases of human infections due to non-cholera vibrios were identified in France by the NRCVC. Five cases of septicemia and/or gastroenteritis and 2 cases of miscellaneous suppurations were caused by V. cholerae non-O1/non-O139. None of the V. cholerae non-O1/non-O139 strains possessed the cholera toxin genes. Three cases of miscellaneous suppurations were caused by V. alginolyticus.
Since 1995, the NRCVC has maintained a voluntary surveillance system for non-cholera Vibrio infections in France. Clinical data, including information about underlying illnesses, and epidemiological data on seafood consumption and exposure to seawater before illness, are systematically collected for each culture-confirmed non-cholera Vibrio infection reported to the NRCVC. In 2005, we summarized data for all the human non-cholera Vibrio isolates reported to the NRCVC in the last ten years, from 1995 to 2004. In total, 93 cases of culture-confirmed Vibrio illnesses were reported.
These Vibrio infections caused several syndromes: gastroenteritis with or without secondary septicemia; wound infections with septicemia; primary septicemia with no evidence of previous gastroenteritis or wound infection; and miscellaneous suppurations. V. cholerae (non-O1/non-O139) was the most frequently reported Vibrio (47.3 % of patients) and was associated with gastroenteritis with or without septicemia, wound infections, and septicemia. None of the V. cholerae (non-O1/non-O139) isolates possessed the cholera toxin genes. V. parahaemolyticus, which was mostly isolated from patients with gastroenteritis, and V. alginolyticus, which was mostly isolated from cases of miscellaneous suppurations, were each isolated from 18.3 % of the patients. Fifteen of the 17 V. parahaemolyticus isolates had either one or two hemolysin (TRH and/or TDH) genes (11 isolates with the tdh gene, 3 isolates with the trh gene, and 1 isolate with both genes). Five isolates belonged to the new pathogenic O3:K6 pandemic clone. Epidemiological data showed that three of the five patients infected with this new O3:K6 clone had eaten local seafood harvested in uncontrolled areas several years apart on the French coast. This provides evidence that the pathogenic V. parahaemolyticus clone O3:K6 is present and suggests that it can persist in the French coastal environment. V. vulnificus was isolated from 11.8 % of patients, mostly from cases of severe wound infection and septicemia. All the V. vulnificus isolates possessed the hly gene.
Sixty-five of the 93 patients (70%) were hospitalized and eight (8.7 %) died. Seven of the eight patients who died presented underlying illnesses. Patients with underlying medical conditions were more likely to develop septicemia. We found that 80.5 % of the patients for whom such information was available were contaminated in France, with the others acquiring the infection abroad. Most infections were associated with eating seafood in the seven days before illness, contact with seawater or handling seafood. Vibrio illnesses were correlated with the warmer months of the year from May to October. It is difficult to monitor Vibrio illnesses because these infections are not notifiable and only cases associated with severe syndromes are reported. Indeed, the official surveillance authority estimated that the number of cases reported by the NRCVC was representative of severe Vibrio infections only. However, despite the small number of Vibrio infections reported in France, these pathogens should be included in microbiological surveillance systems, because of the increase in the number of people susceptible to infection, the severity of infections and the recent introduction and probable implantation of a pandemic clone of V. parahaemolyticus in France.
The NRCVC is also responsible for identifying non-cholera vibrio strains isolated from seafood imported into France. In 2005, V. cholerae, V. parahaemolyticus and V. alginolyticus were the most frequently identified species among the 172 strains studied. All seafood isolates of V. vulnificus had the hly gene, indicating that all strains of this species are potentially pathogenic. By contrast, only a very small percentage (<1%) of strains of V. parahaemolyticus isolated from imported seafood had a hemolysin gene. None of the V. cholerae non-O1/non-O139 strains isolated from seafood samples possessed the cholera toxin genes.
In collaboration with the Institut Pasteur de Madagascar and the Institut National de Recherche Halieutique in Morocco, we have studied 173 non cholera vibrios strains isolated from seafood of from seawater.
The NRCVC has participated in postgraduate training programs and its members have given several lectures on cholera and other vibrio infections. The NRCVC is also involved in the training of microbiologists and clinicians. We have run courses in Morocco and Cameroon on classical and molecular methods for studying cholera and non-cholera vibrios, for microbiologists, technicians and students from several countries (Algeria, Bangladesh, Cameroon, Central African Republic, Ivory Coast, Democratic Republic of Congo, France, Madagascar, Mali, Mauritania, Mexico, Morocco, Senegal, Tunisia).
Rapid diagnostic test for cholera : the kit includes everything required for the field diagnosis of cholera: individually packed dipsticks in aluminum pouches; test tubes for the sample; disposable plastic droppers for the transfer of stool samples to test tubes; a test-tube stand.
Keywords: cholera, vibrios, vaccination, conjugate, diagnosis, molecular epidemiology, seafood, public health