Michel Tibayrenc
Centre d'Etudes sur le Polymorphisme des Microorganismes (CEPM), UMR CNRS/ORSTOM 9926, ORSTOM, BP 5045, 34032 Montpellier Cedex 01, France
Strain typing
Molecular taxonomy
References
In the present paper, rather than proposing an extensive survey of the literature published on the topic, I will present the approach developped by my group to address the general problem of molecular epidemiology of pathogenic microorganisms, and I will take the illustrative examples of 2 salivarian trypanosomes, Trypanosoma brucei, the agent of sleeping sickness, and Trypanosoma congolense, the agent of cattle trypanosomiasis, both transmitted by tsetse flies in Africa. I will especially try to emphasize some traps hidden in the molecular epidemiology approach.
Our approach is based on the 2 following principles:
(i) integration between basic science (evolutionary genetics) and applied research (molecular epidemiology);(ii) comparative approach, with analysis by the same team, with the same technical and theoretical tools, of various pathogens, bacteria, yeasts, parasitic protozoa. As an example, in our laboratory, we have presently programmes on Trypanosoma cruzi, T. brucei, T. congolense, Leishmania spp., Candida albicans and Mycobacterium tuberculosis. The interest of the compared approach is to evidence the general laws that govern the genetic diversity of microbes, and at the same time, to see clearly the pecularities of each species. An immediate pay-off of this approach is to design versatile and polyvalent molecular tools for identification, useable for any kind of microorganism. These principles have been described in recent review papers (Tibayrenc, 1995, 1996).
The 3 main applications of evolutionary genetics in the field of applied microbiology are: (1) Molecular typing of strains, for epidemiological follow-up (2) Taxonomy/classification (3) Analysis of the relationships between genetic diversity of the microorganisms and their relevant medical properties such as virulence or resistance to drugs. Only the 2 first points will be developed here, for they have direct implications in molecular epidemiology.
Many teams rely on an empirical approach to type strains of pathogens. They analyse visually the profiles generated by the molecular tools they use, and consider that the stocks that appear identical pertain to the same "strain ", and therefore, are involved in the same epidemic chain. This approach can be misleading for 2 reasons:
(i) the "identity" of strains in this case is highly dependent upon the resolution power of the molecular tools employed. As an example, when random primer amplified polymorphic DNA (RAPD, a special kind of PCR technique relying on arbitrarily designed decameric primers) is used, a set of stocks that appear identical with 4 primers will prove to be heterogeneous when 20 primers are used. Considering that these stocks pertain to the same " strain " is therefore misleading. Actually, it is practically impossible to prove the identity of 2 genotypes, unless they are fully sequenced. One has to make do with sufficient presumptions, based on the use of genetic tools that have a fair level of resolution (in my laboratory, we use routinely up to 22 isoenzyme loci + 25 RAPD primers). This problem of the level of resolution is tightly linked to the " molecular clock " (fastness of evolution) of the genetic tool employed.(ii) Strain typing implicitly relies on the assumption that the genotypes of the species under study are sufficiently stable in space and time to be reliably characterized. This could be untrue if the species considered undergoes regular genetic recombination. In this last case, the genotypes of this species are ephemeral individual variants whose genetic make-up is reshuffled regularly. "Strains" of such species are therefore improper as epidemiological markers. Such "sexual" microbial species include for example Plasmodium falciparum, the agent of malaria and Neisseria gonorrhoeae, a bacterium responsible for gonorrhoea. On the contrary, if genetic recombination is rare or absent, the species is clonal, and its strains have a stable genotype. They reproduce themselves like "genetic photocopies", and can be reliably used as epidemiological markers.
Trypanosoma cruzi, the agent of Chagas disease, and the bacteria pertaining tho the genus Salmonella, seem to be typically clonal. The debate clonality/sexuality is therefore crucial for the problem of strain typing and molecular epidemiology. The proper way to address the question is a population genetic approach based on the analysis of linkage disequilibrium (nonrandom association of genotypes observed at different loci). If the species is sexual, little or no linkage is observed. Strong linkage is considered circumstancial evidence for clonality (see Tibayrenc, 1995, 1996). The case of Trypanosoma brucei still is under debate. There is little doubt that this species can propagate itself clonally, but mating has been obtained experimentally (Jenny et al., 1986). Moreover, some authors (Maynard Smith et al., 1993; Hide et al., 1994) propose that T. brucei is actually a sexual species, at least in some transmission cycles, and that the clones evidenced by population genetic analyses (Mathieu-Daude & Tibayrenc, 1992) are " epidemic clones " which last at best a few years.
If this is verified, the interest of T. brucei strains as epidemiological markers is strongly disminished. The intense debate on T. brucei genotype stability (see also Mathieu-Daude et al., 1996; Stevens & Tibayrenc, 1995, 1996) illustrates the traps hidden beyond the apparently simple matter of strain identification, and the extreme conceptual difficulties for addressing the question.
Trypanosoma congolense exhibits strongs indications of linkage disequilibrium (Tibayrenc et al., 1990; Sidibe, 1996) and is probably a clonal species. Nevertheless, the populations analyzed until now have not been collected in rigorous sympatric conditions (sympatric in a given locality). This still has to be done before we can accept that T. congolense strains are safely useable for epidemiological tracking. Indeed fair sympatric samples are necessary for a reliable population genetic analysis.
Evolutionary genetics can be useful for perfecting taxonomy in 2 circumstances:
(i) it can show to which extent the presently described species correspond to real genetic entities;(ii) within the presently-described species, it can evidence hidden genetic subdivisions that deserve to be described as separate taxa. In the two cases, the job of the evolutionist is the same: to decide whether the categories he is considering correspond to actual phylogenetic subdivisions (clades). The implications for molecular epidemiology are considerable. Indeed this conditions the possibility to use specific molecular tools (probes, PCR diagnoses), able to identify all the isolates of the species (or of a given subdivision of the species), and only them. If a species is polyphyletic, if it is actually an artificial taxon composed of different clades that have little to do together, it will be impossible to design molecular tools that will act as a "genetic common denominator", and will be able to identify all the isolates of this species, and only them.
The case of Trypanosoma brucei:
(i) there is little doubt that the whole species T. brucei corresponds to a separate clade, definitely distinct from other species. T. brucei is monophyletic, and in a phylogenetic point of view, is a "good" species.
The practical consequence is that it is possible to design molecular tools specific of this species.
(ii) Within T. brucei sensu lato, it is not totally clear whether the three "subspecies" classically described (T. brucei congolense, the agent of West african human trypanosomiasis, T. brucei rhodesiense, the agent of East African Human trypanosomiasis, and T. brucei brucei, the agent of cattle trypanosomiasis) correspond to real clades. This seems to be the case (Mathieu-Daude & Tibayrenc, 1994; Mathieu-Daude et al., 1995) for the group of genotypes referred to as "Trypanosoma brucei gambiense group 1" (Gibson, 1986). It has been consequently possible to design a kDNA probe specific for this monophyletic subdivision of T. brucei s.l. (Mathieu-Daude et al., 1994). Similarly, Hide et al. (1996) have proposed that T. brucei rhodesiense is composed of a set of genotypes distinct from any other T. brucei s.l. It would therefore correspond to a monophyletic group.
The case of Trypanosoma congolense
This parasite shows possibly the opposite picture. The PhD research by Sidibe (1996) in our group shows that:
(i) in confirmation of previous studies (Gashumba, Baker & Godfrey, 1988), 3 groups of parasites, corresponding to monophyletic entities subdivide the species T. congolense. These 3 subdivisions (referred to as: " Savannah "," Killifi" and " Forest", Gashumba, Baker & Godfrey, op. cit.) can be safely identified by several isoenzyme and RAPD characters (Sidibe, 1996).(ii) with extensive genetic characterization (18 isoenzyme loci and 23 RAPD primers), the whole T. congolense does not appear monophyletic (Sidibe, op. cit.). One cannot separate the whole group from related species, such as T. simiae and T. brucei. It is impossible to decide whether this is due to either real polyphyletism, or to the fact that the upper level of resolution of isoenzymes and RAPDs is saturated at this high level of phylogenetic divergence (Tibayrenc, 1995, 1996). If it is confirmed that T. congolense is polyphyletic (and hence, is an artificial species from a phylogenetic point of view), the practical implication is that it will be impossible to design molecular tools of identification specific for the whole species.
My hope through this review paper is to call the attention of epidemiologists and clinicians to the unsuspected difficulties of molecular epidemiology, and to dissuade the potential molecular epidemiologists from trusting empirical, intuitives approaches.
Gashumba, J.K., Baker, R.D. & Godfrey, D.G. 1988. Trypanosoma congolense: the distribution of enzymic variants in East and West Africa. Parasitology, 96: 475-487.
Gibson, W.C. 1986. Will the real Trypanosoma brucei gambiense please stand up? Parasitol. Today, 2: 255-257.
Hide, G., Tait, A., Maudlin, I. & Welburn, S.C. 1996. The origins, dynamics and generation of Trypanosoma brucei rhodesiense epidemics in East Africa. Parasitol. Today, 12: 50-55.
Hide, G., Tait, A., Maudlin, I. and Welburn, S.C., 1994. Epidemiological relationships of Trypanosoma brucei stocks from South East Uganda: Evidence for different population structures in human infective and non-human infective isolates. Parasitology 109: 95-111.
Jenni, L., Marti, S., Schweizer, J., Betschart, B., Le Page, R.W.F., Wells, J.M., Tait, A., Paindavoine, P., Pays, E. and Steinert, M., 1986. Hybrid formation between African trypanosomes during cyclical transmission. Nature, 322: 173-175.
Mathieu-Daude, F. and Tibayrenc, M., 1994. Isozyme variability of Trypanosoma brucei s.l.: genetical, taxonomical and epidemiological significance. Exp. Parasitol., 78: 1-19.
Mathieu-Daude, F., Bicart-See, A., Bosseno, M.F., Breniere, S.F. and Tibayrenc, M., 1994. Identification of Trypanosoma brucei gambiense group I. by a specific kinetoplast DNA probe. Am. J. Trop. Med. Hyg, 50 (1): 13-19.
Mathieu-Daude, F., Stevens, J., Welsh, J., Tibayrenc, M. and McClelland, M., 1995. Genetic diversity and population structure of Trypanosoma brucei: clonality versus sexuality. Mol. Biochem. Parasitol, 72: 89-101
Sidibe, I., 1996. Variabilite genetique de Trypanosoma congolense, agent de la trypanosomose animale: implications taxonomiques et epidemiologiques. PhD dissertation, University of Montpellier, France.
Stevens, J.R. and Tibayrenc, M., 1995. Detection of linkage disequilibrium in Trypanosoma brucei isolated from tsetse flies and characterized by RAPD analysis and isoenzymes. Parasitology, 110; 181-186.
Stevens, J.R. and Tibayrenc, M., 1996. Trypanosoma brucei s.l.: evolution, linkage and the clonality debate. Parasitology 112: 481-488.
Tibayrenc, M., 1995. Population Genetics of Parasitic Protozoa and other Microorganisms. Advances in Parasitology (Eds. Baker, J.R., Muller, R. and Rollinson, D.): 36: 47-115.
Tibayrenc, M., 1996. Towards a unified evolutionary genetics of microorganisms. Ann. Rev. Microbiol. 50: 401-429.
Tibayrenc, M., Kjellberg, F. and Ayala, F.J., 1990. A clonal theory of parasitic protozoa: the population structure of Entamoeba, Giardia, Leishmania, Naegleria, Plasmodium, Trichomonas and Trypanosoma, and its medical and taxonomical consequences. Proc. Nat. Acad. Sci. USA: 87: 2414-2418.
