Adaptation to environmental stress in different life stages of <i>Drosophila melanogaster</i>

Even if Charles Darwin was not the first to elaborate a theory about evolutionary biology, he was the first to elaborate a truly scientific theory. Effectively the theory of natural selection expressed in On the Origin of Species by Means of Natural Selection (Darwin, 1859) is based on numerous observations, scrupulously analysed and discussed. Historically, intra- and interspecific competition have been considered as the main factors in adaptation and the evolutionnary process. However, recent insights show that environmental stresses have a large impact on the genetic structure of populations, and on the evolution of biological systems (Calow and Berry, 1989; Hoffmann and Parsons, 1991; Bijlsma and Loeschcke, 1997a). An environmental stress can be defined as an environmental factor that reduces Darwinian fitness. The occurrence of this kind of event with high selection pressure may be rare, but it has a large impact on the evolutionary processes, from the molecular level to the ecosystem level. Environmental stress may even play a role in speciation processes (Parsons, 1997; Sheldon, 1997). Holometabolous insects present a complete metamorphosis with a larval stage entirely different from the adult stage. The larval stages are specialized in growth while the adult stage is specialized in dispersal and reproduction. Larvae and adults generally have a different shape, may live in different environments and may even have a different alimentary regime. A similar environmental stress can affect each life stage differently, and the adaptative response to a particular stress may be dependent on or even specific to a particular life stage (Loeschcke and Krebs, 1996). Furthermore, an adaptative response to a single environmental stress in a particular life stage may have an impact on other life stages. The aim of this thesis is to analyse the genetic effects on the different life stages of an environmental stress on a holometabolous insect population. Alcohol resistance in Drosophila melanogaster is well documented (van Delden and Kamping, 1997) and provides an excellent model system to study environmental stress. The feeding substrate in the wild of this species is constituted by decaying fruits which can contain various ethanol concentrations due to fermentation. The presence of toxic levels of ethanol in the medium can be considered as an environmental stress. On the one hand, adult feeding behaviour and female oviposition behaviour imply that adults are potentially in contact with high ethanol levels. On the other hands, larvae feed and grow in the substrate and cannot escape the environment, whatever the alcohol concentration may be. Environmental ethanol may affect both juvenile and adult stages. However, feeding and oviposition behaviour do not necessarily involve a long exposure to ethanol stress for adults, and the selection pressure of ethanol is probably lower for the adults than for the larvae (Boulétreau and David, 1981). Alcohol dehydrogenase (ADH: EC 1.1.1.1) is a key enzyme in the metabolic pathway of alcohols in D. melanogaster and is essential for survival in ethanol environments, as shown by the sensitivity of Adh null mutants (David et al., 1976). The two common variants, AdhS and AdhF, differing in one nucleotide, are the cause of considerable differences in enzyme activity: The in vitro ADH activity AdhFF flies is generally about three times higher than AdhSS flies, while heterozygotes display intermediate activity (van Delden, 1982; Chambers, 1988, 1991; Heinstra, 1993). Generally alcohol tolerance is positively correlated with ADH activity, AdhFF individuals showing higher resistance to ethanol stress than AdhSS individuals. However, ADH activity and alcohol tolerance show considerable variation in natural populations, even within a single Adh genotype. In D. melanogaster, alcohol tolerance involves several mechanisms. Some of them are probably different according to the life stage, and even specific to a particular life stage. In order to investigate these mechanisms, we have applied three different procedures of selection for increased resistance to environmental ethanol according to the life stage. In order to discriminate between the involvement of both Adh genotypes, two lines, one homozygous for AdhS and the other homozygous for AdhF, were both selected to increase specifically adult ethanol tolerance (ADU-SS and ADU-FF), or larval ethanol tolerance (LAR-SS and LAR-FF). The third selection line was selected for increased ethanol tolerance during the whole life (WHO-SS and WHO-FF) [Chapter 2]. In the first selection procedure (ADU), individuals were selected only at the adult stage, while during the juvenile stages the individuals were never in contact with alcohol. Flies were allowed to mate and females to lay eggs on standard medium. Larvae were grown in medium without alcohol, and the emerging adults were transferred and kept on normal food for one week. Then, flies were transferred into bottles with food supplemented with ethanol. When approximately a quarter of the flies were dead, the survivors were transferred into bottles with standard medium for an egg-laying period of 24 hours to start a new generation. In the second selection procedure (LAR), on the contrary, the juvenile stages were fed and grown in an alcohol-supplemented medium, and the adults were kept on standard medium without ethanol. Newly emerged flies were kept on standard medium for one week. Then, one-week-old females were allowed to lay eggs during a period of four hours, thereafter the eggs were transferred to five bottles containing ethanol food. Larvae developed in this medium, and emerging adults were daily transferred into new bottles with fresh standard food, and kept on this medium for one week before starting the next generation. In the third selection procedure (WHO), the individuals were continuously kept on ethanol medium during their complete life cycle. Females laid eggs on ethanol food, larvae grew in this medium and emerging adults were kept in the bottles for one week. Then, the flies were transferred to bottles with ethanol-supplemented food to allow the females to lay eggs to initiate a new generation. The two initial populations were kept as control populations on regular food without ethanol during the whole life cycle (CON-SS and CON-FF). After 20 (LAR and WHO selection procedures) or 25 (ADU selection procedure) generations of selection, direct responses (adult survival and egg-to-adult survival on ethanol medium) and indirect responses (ADH activity, body weight, protein content and developmental time) were investigated in larvae and adults of all selected and control lines [Chapter 2]. A significant increase in adult survival was observed in the two lines selected for increased adult resistance, ADU-SS and ADU-FF, but in the lines selected for increased juvenile resistance, LAR-SS and LAR-FF, the increase in adult survival was not significant. For the lines selected for increased ethanol resistance during the whole life, WHO-SS and WHO-FF, the increase in adult resistance was significant only in AdhFF males. On the contrary, egg-to-adult survival increased significantly in LAR and WHO selection lines for both Adh genotypes, but not in the ADU-SS and ADU-FF lines. The increase in alcohol tolerance was thus specific for the life stage according to the selection procedure. These direct responses to selection indicate that effectively an environmental stress may act specifically in a particular life stage. Indirect responses were also investigated after 20 or 25 generations of selection. An increase in adult ADH activity seems to be related with the increase in adult resistance to ethanol in the ADU lines, while larval ADH activity was similar in the LAR lines and the control lines (CON). An increase in body weight, linked to an increase in development time, is observed in the AdhFF selected lines, and plays a role in the increase in both adult and juvenile resistance. These responses indicate that ethanol resistance is a complex trait involving probably several loci. The results on ADH activity confirm that the responses to environmental ethanol are different according to the life stage. An increase in constitutive ADH is linked to the increase in adult tolerance while this is not the case for juvenile tolerance to ethanol. The significant increase in body weight underlines the fact that an environmental stress can have various responses, not always specific to the stress, and can have various effects on the evolutionary processes in a population. The role of ADH in the specific increase in alcohol tolerance according to the life stage was investigated more deeply after 40 (LAR and WHO) or 45 (ADU) generations of selection [Chapter 3]. Results of juvenile and adult survival on ethanol medium confirm the results observed after 20 or 25 generations of selection. Selecting for increased ethanol resistance leads to a positive response in adults, irrespective of the selection regime. For egg-to-adult survival, however, a positive and significant response was observed only when the larval stage is included in the selection regime. Selection in the larval stage then extends to the adult stage. Many developmental processes are in operation during the larval stage or will be implemented in later stages. It is likely that similar adaptations of the metabolic pathway constitute ethanol resistance in adults and larvae. However, the additional developmental processes in the larval stage make adaptative changes in additional pathways necessary. This would explain the fact that resistance obtained in the larval stage, including metabolic and additional pathways, extends to the adult stage. The increase in adult survival in ADU-SS and ADU-FF lines was accompanied by an increase in adult ADH activity. On the contrary the increase in egg-to-adult survival in LAR-SS and LAR-FF was not accompanied by a significant increase in larval ADH activity. It is argued that the role of ADH in the increase in ethanol tolerance is dependent on the life stage and on the Adh genotype. A positive correlation between ADH protein abundance and ethanol tolerance for adults but not for larvae has already been reported by Geer et al. (1993). Adaptation to environmental ethanol is correlated with higher ADH activity. However in larvae this higher activity may be in conflict with other pathways. It is possible that ADH is involved in other metabolic pathways than ethanol degradation only, as suggested by Oppentocht et al. (2002). Presence of alcohol in the feeding medium induces an increase in Adh expression and ADH activity in larvae, and, to a lesser extent, in adults [Chapter 4]. The increase in adult resistance to ethanol was accompanied by an increase in adult ADH activity in AdhFF selected lines on ethanol medium. On the contrary in AdhSS selected lines, adult ADH activity did not increase after one day on ethanol medium. The increase in larval ADH activity on ethanol medium was also more pronounced in AdhFF selected lines compared to the control line, while the increase was identical for LAR-SS and CON-SS. Concerning ADH activity, responses to selection appear to be different according to the Adh genotype. Previous research has demonstrated that Adh mRNA levels are not necessarily associated with higher ADH activity or, indeed, alcohol tolerance (Choudhary and Laurie, 1991; Laurie et al., 1991). Moreover the regulation of ADH activity can be effectuated at different levels (Stam and Laurie, 1996; Carlini, 2004). However the differences observed in the results for the two Adh genotypes confirm the idea that the adaptation to high levels of environmental ethanol involves probably other physiological systems. For two selected lines ADU-FF and LAR-SS, reciprocal crosses between selected and control lines (respectively CON-FF and CON-SS) were performed [Chapter 5]. The offspring was tested for adult ethanol resistance (ADU-FF x CON-FF) or juvenile ethanol resistance (LAR-SS x CON-SS). In both cases, results show no X-chromosome effect. In the first case, results suggest a general codominant effect for the increase in adult ethanol resistance, while in the second case results suggest a dominant effect for the increase in juvenile resistance. In another experiment crosses with a balanced marker line were performed to construct genotypes in order to analyse the effects of X, second and third chromosomes of the two selected lines on the increase in adult and juvenile ethanol tolerance. In all cases, the X chromosome has no effect, while the second chromosome always has a significant effect on adult and on juvenile tolerance to ethanol. The third chromosome from LAR-SS has no effect on ethanol tolerance, but the third chromosome from ADU-FF has a significant effect on juvenile tolerance and on male adult tolerance. The analysis of the results of different selection procedures according to the life stage points to complex interactions between ethanol stress, life stage, sex and genotype. It appears clearly that selection for increased alcohol tolerance is not only linked with modifications of the Adh gene but has also effects on other traits like body weight and developmental time. It is concluded that several loci are involved in the increase in alcohol tolerance, with a relative specificity according to the life stage. Furthermore, preliminary results about acetic acid resistance show that a second environmental stress, often linked with alcohol, involves probably other loci. The results of the thesis confirm the importance of abiotic environmental stresses in the evolutionary processes. They further show that the different life stages of a holometabolous insect may react quite differently on specific environmental stresses and involve distinct genetic mechanisms