Effect of inbreeding on trait stability in Steinernema feltiae

Transcription

1 University of Ghent Faculty of Science Department of Biology Academic Year Effect of inbreeding on trait stability in Steinernema feltiae Sevgi TÜRKÖZ Promoter: Prof. Ralf-Udo EHLERS Supervisor: Dr. Olaf STRAUCH Thesis submitted to obtain the degree of European Master of Science in Nematology

2 Effect of inbreeding on trait stability in Steinernema feltiae Sevgi TÜRKÖZ Department of Biology, Ghent University, K.L. Ledeganckstraat 35, B-9000 Ghent, Belgium Summary- Progress in genetic improvement of beneficial traits of entomopathogenic nematodes (Steinernema and Heterorhabditis) by hybridisation and genetic selection has often been lost again once the selection pressure was released during mass production. Recently, a method was developed for Heterorhabditis bacteriophora to stabilize traits by selection of improved inbred lines. Culturing this nematode in liquid media results in highly homozygous inbred lines, because under these conditions the nematode reproduces exclusively through selfing. This approach is not possible for Steinernema spp. because they are amphimicitic. Therefore, the main objective of this study was to develop a procedure to produce inbred lines of a desiccation tolerant hybrid strain of Steinernema feltiae (HYB13A) and check how inbreeding influences beneficial traits like desiccation tolerance and the virulence. Results were compared with a commercial strain (Com). Inbred lines were produced by culturing single gravid females isolated from infected insects on agar plates seeded with the symbiotic bacterium X. bovienii and continuously transferring those single gravid females up to eight inbreeding steps. Survival of inbred lines was low due to contaminations carried over from insect cadavers. An antibiotic treatment of the females could not solve the problems. Culturing nematodes in monoxenic culture was more successful, mean inbreeding steps until lines were lost was 4.5 whereas inbreeding with females originating from insects (in vivo) with or without antibiotic treatment provide mean inbreeding step of 2.5. The desiccation tolerance of the S. feltiae strains was measured by producing dehydrating conditions using different concentrations of polyethleneglycol 600 and 2

3 the water activity (a w ) was measured. Before inbreeding the mean tolerated water activity survived by 50% of the population (WA 50 ) measured after a 48h adaptation to desiccation stress was for strain Com and for the HYB strain, which had been selected for enhanced tolerance, was assessed without a significant difference between the strains. Significantly more tolerant were S. riobrave strain Sr Lin4 (a w -value= 0.73) and S. yirgalemense (a w -value= 0.769). Only few inbred lines could be monitored for the development of their desiccation tolerance during inbreeding because many were lost with increasing inbreeding steps. For HYB strain (Line 16 and 17) WA 50 significantly decreased from 0.83 to and 0.658, respectively. Com strain line 20 also dropped from 0.91 to There was a significant correlation between the desiccation tolerance and the inbreeding step for the lines that can be followed for the continuous inbreeding step. The HYB strain was significantly more tolerant than the Com strain for the mean tolerated water activity also when exposed to stress without prior adaptation of the population. Inbred lines with selection pressure was significantly higher tolerant than those without selection pressure. This significance was only for the most tolerant 10% individuals of HYB strain which dropped from 0.9 to for the non-adapted populations. The virulence seemed to be negatively affected for the HYB strain. The investigation was the first attempt to try to stabilize beneficial traits of S. feltiae through production of inbred lines. Keywords- Biological control, Steinernema feltiae, desiccation tolerance, inbreeding, selection Entomopathogenic nematodes (EPNs), Steinernema (Travassos 1927) and Heterorhabditis (Poinar 1976) are associated with the symbiotic bacteria Xenorhabdus (Thomas & Poinar, 1983) and Photorhabdus (Akhurst et al., 1996), respectively. Although these bacterial symbionts belong to the same family (Enterobacteriaceae), they form a different phylogenetic group (Ehlers et al., 1988). EPNs are safe biological control agents and 3

4 have been successfully used against soil borne insects in ornamental plants, turf, mushrooms and strawberries (Kaya and Gaugler, 1993; Ehlers, 1996). Like the other orders of Rhabditida, formation of dauer (enduring) juveniles (DJs) is also associated with Steinernema and Heterhorhabditis spp. They are adapted to long term survival and carry between 0 and 250 cells of symbionts in the anterior region of their intestine (Spiridonov et al., 1991; Glazer, 1996). After penetration of DJs into an insect, the symbiotic bacteria are released into the haemolymph. These bacteria propagate and produce toxins (Dowds and Peters, 2002) and other metabolites (Websteri et al., 2002), which help to suppress the defence mechanism of insect hosts, which usually die within 48 hours after the invasion of the nematode. Heterorhabditis spp. can not kill the insect in the absence of its symbiont P. luminescens (Han and Ehlers, 2000). In the case of Steinernema spp., it could produce toxins that facilitate the pathogenicity (Burman, 1982; Ehlers et al., 1997). The usage of EPN is safe for the environment and has no detrimental effects on mammals (Ehlers and Peters, 1995; Grewal et al., 2005). In many countries, they are exempted from registration requirements (Ehlers, 2011). They can be produced commercially in solid and liquid culture systems (Ehlers, 2001). However, one of the main constraints of EPN is their limited shelf life in transportation and storage. In order to keep the quality and viability of the nematodes, they are mixed with inert clay minerals and kept at low temperature (Strauch et al., 2000). Another alternative is to desiccate the DJs at a water activity (a w -value) of 0.97 (Grewal and Peters, 2005). The vapour pressure of a liquid divided by that of pure water at the same temperature is referred to as water activity. It indicates the intensity at which water is bound to solid compounds in a sample (Anbesse et al., 2012a). It ranges between 0 to 1. The lower the value of the water activity (a w -value), the harder the removal of the water, hence the higher the desiccation tolerance of an organism. Desiccation is regarded as a major way of getting storage stability of EPN since it induces anhydrobiosis 4

5 (a reversible physiologically arrested state of dormancy) (Barret, 1991). When a nematode enters anhydrobiosis, it accumulates trehalose and glycerol which can replace the water in biological membranes (Womersley, 1990; Solomon et al., 1999). However, lowering a w - values is also still not enough since it can not stop the growth for bacteria and fungi. In order to find solutions, researchers were mainly focused on enhancement of the beneficial traits with genetic selection and domestication of nematodes by improving these traits which are favorable for EPN commercialization. The most applicable method for the enhancement of beneficial traits is selective breeding (Burnell, 2002). Success of breeding is affected by the heritability, selection pressure, and variability within population (Strauch et al., 2004). Increased desiccation tolerance of EPN strains have been achieved by several researchers. This researches was mainly focused on H. bacteriophora (Strauch et al., 2004; Mukuka et al., 2010), which was enhanced in desiccation tolerance. The desiccation tolerance of S. feltiae was improved by Nimkingrat et al. (2013a) with hybridization of the most tolerant strains and further exposing them to genetic selection over six cycles. Selection was done for tolerance including an adaptation phase prior to expose to higher stress conditions and the strain was named as HYB13. After six selection cycles, the mean tolerated water activity (WA 50 ) dropped from to 0.69 for adapted populations and from to for tolerance to immediate desiccation stress without any adaptation of the nematode population to lower water activity. However, the HYB strain lost the gained desiccation tolerance after few propagation cycles through Galleria. The same problem was reported for H. bacteriophora (Anbesse et al., 2012a). Before introducing such strains for commercial use, scientists should prove whether the traits are stable also during mass production. It should also be tested whether other beneficial traits like host finding, infectivity, reproduction potential, adaptability to liquid culture conditions (Ehlers, 2001) and storage potential (Strauch et al., 2000) are maintained during selective breeding. These traits may decrease or disappear after 5

6 selection processes like cross-breeding, inbreeding, unintentional selection and drifting (Hoy, 1985; Roush, 1990; Hopper et al., 1993; Grewal et al., 2006). Developing methodologies to prevent or to stabilize the trait loss in EPN is therefore a critical point that should be further improved. For instance, usage of inbred lines could be a possibility to prevent trait deterioration (Dolgin et al., 2007). Inbreeding is defined as reproduction of parents that are genetically closely related, e.g. crossing of brother and sisters. It enhances the homozygosity; therefore, increases the chances of offspring being affected by recessive and deleterious traits (Nabulsi et al., 2003). Inbred lines are formed as a result of crossing offspring of the same female followed by continuous selfing or sibling mating in order to create new inbred line with the mosaic of the parental genome (Falconer and Mackay, 1996). To be able to get pure inbred lines with homozygosity close to 99%, crossing should be done more than seven generations if it starts from hybrids. It has been reported that inbred lines are more stable than populations originating from in vivo production and some of the inbred lines even have better traits than their parental populations (Bai et al., 2005). Another approach to limit trade-off effects is cryopreservation which helps to maintain the genetic diversity (Bai et al., 2004). However, inbreeding can also result in inbreeding depression and even extinction of inbred populations (Dolgin et al., 2007). As reported for Caenorhabditis elegans inbred depression is mostly low in highly selfing species also such as Heterorhabditis species due to the reproduction through hermaphrodites in the F1 populations. Another important finding is that inbred line of C. elegans performed better than crosses between strains in life history trait analysis even showed outbreeding depression (Dolgin et al., 2007). Apart from H. bacteriophora, S. feltiae and S. carpocapsae have also been produced in large scales in bioreactors (Ehlers, 2001). S. feltiae has a symbiotic relationship with the bacterium Xenorhabdus bovienii Thomas and Poinar (Proteobacteria: Enterobacteriaceae). 6

7 The mass production is accomplished in bioreactors at industrial scale (Ehlers, 2007). S. feltiae reproduces by crossing of males and females (amphimixis). The DJ of S. feltiae develops to amphimictic adults and their offspring either develop to DJ or F1 generation (Hirao et al., 2010). Development of F2 often does not occur due to the depletion of food sources, thus DJs are formed (Ehlers, 1996). In Heterorhabditis spp., reproduction in liquid culture is only by self-fertilising hermaphrodites due to the copulation behaviour of the nematode. Under liquid culture conditions the amphimictic adults cannot mate; hence, allowing automatic selection for inbred lines (Strauch et al., 1994). As it was observed in the study of Anbesse et al. (2012a) for H. bacteriophora, increased desiccation tolerance obtained through genetic selection was maintained at higher level after the release of selection pressure during seven reproduction cycles in in vitro production whereas it was lost during in vivo production allowing cross fertilization. Furthermore, the mean desiccation tolerance was significantly increased from a w of 0.81 to 0.76 when the homozygous inbred lines were propagated in vivo, allowing the cross breeding of homozygous lines (Anbesse et al., 2012b). The trait values in progeny for desiccation tolerance increased over the level recorded for inbred parental lines when allowed to crossing of inbred lines of H. bacteriophora from liquid culture. This heterosis effect in EPN was first investigated in this study and considered as a big step for the improvement of beneficial traits in the commercial production of EPN. This approach is possible also with Steinernema spp. although their amphimictic mode of reproduction and crossing behavior under liquid culture, but would need a much more laborious approach as selection of tolerant inbred lines cannot just been done by putting selection pressure on population during liquid culture as successfully done by Anbesse et al. (2012a). The aim of this study was (1) to develop a method to produce inbred lines also with the nematode S. feltiae that is reproducing by amphimixis and (2) to find out how beneficial 7

8 traits like desiccation tolerance develop during the inbreeding process, whether traits change and whether inbred lines are even lost due to inbred depression. Materials and Methods NEMATODE STRAINS AND INSECT CULTURE The nematode species and strains used in this study are listed in Table 1. The strains HYB13A (Nimkingrat et al., 2013a) and Com (Peters, 1994) are the hybrid strains from several natural isolates of S. feltiae. The natural isolates used for hybridisation are listed in Table 2 and 3, respectively. HYB13A was obtained after crossing six strains and then exposing the hybrid to six selection cycles including an adaptation phase to desiccation stress. Com strain was supplied by e-nema GmbH. Those two hybrid strains were the major strains used for inbreeding experiments. Before and after inbreeding, they were checked for the desiccation tolerance. On the other hand, S. yirgalemense (Sy 157-c) and 3 different strains of S. riobrave (Sr Tp, Sr 3-7 and Sr Lin4) were used for the desiccation tolerance experiment for comparison with the hybrid strains of S. feltiae. Table 1. Strains of Steinernema feltiae (Sf), S.riobrave (Sr) and S. yirgalemense (Sy) with their geographical origin Strains Origin Sf HYB13A Hybrid (Nimkingrat et al., 2011) Sr Tp Sr 3-7 Sr Lin 4 Sy 157-c Texas Texas Florida South Africa Sf Com EN02 (Peters, 1994) 8

9 Strain HYB13A was taken from liquid nitrogen and thawed in warm Ringer s solution (9.0 g NaCl; 0.42 g KCl, 0.37 g CaCl 2 2H 2 O ; 0.2 g NaHCO 3; ad 1 l aqua dest.) for 24 h to get it activated. All nematode strains were propagated in vivo using the last instars larvae of Galleria mellonella (Lepidoptera: Pyralidae) as described by Kaya & Stock (1997). Different batches of Galleria larvae (each batch with 5 larvae) were inoculated with DJs/ larva and kept at 25 C in the dark. After 3-4 days of inoculation, dead cadavers of Galleria larvae were transferred to White traps in order to allow the emergence of DJs. Freshly emerged DJs were collected from white traps within two weeks and stored in Ringer s solution at 14 C and used within three to four days for desiccation tolerance experiments. Galleria mellonella larvae were reared on artificial medium (22% corn-groats, 22% wheat-flour, 11% honey, 11% glycerol, 5.5% yeast powder, 17.5 % bee wax) at 35 C in the laboratory of e-nema GmbH. The experiment was conducted with at least three replications by using different batches of G. mellonella for each replicate. Table 2. Strains used to produce the hybrid (HYB13A) of Steinernema feltiae (Nimkingrat et al., 2013a) Strain Code CR1 NEP1 RUS1 TUR1 ITA1 NL1 Origin Czech Republic Nepal Russia Turkey Italy The Netherlands 9

10 Table 3. Strain designation and their origin of Steinernema feltiae strains used to produce the hybrid Com (Peters, 1994) Strain Origin Strain Origin code code Imm1 Immenstedt, Germany (1992) S-184 Italy Kiel Kleve, Germany (1992) Skr Seekrug, Germany K1e2 Kleve, Germany (1992) SF-S22 Helsinki,Finland K1e3 Kleve, Germany (1992) SF3 SegebergerForst, Germany N-V17E Vestfold,Norway(1989) Sma Südermarsch, Germany OBSIII Ens, The Netherlands Umea Sweden (1981) Oro4 Ostrohe, Germany (1992) Wo3 Szent-TarnasYugoslawia ASSESSMENT OF THE DESICCATION TOLERANCE The desiccation assessment was conducted in two phases including adaptation and without adaptation to desiccation stress according to the protocol developed by Strauch et al. (2004). In order to create the desiccation stress, nematodes were exposed to different concentration of polyethyleneglycol 600 (PEG) solutions which is viscose, non-ionic and nontoxic to DJs. Different ranges of dilutions were prepared with several water activities by mixing the PEG solution with water. A water activity meter (Aqua Lab Model CX-2, USA) was used to measure the water activity. For each strain, ranges of a w -values were tested in order to assess the mean tolerated water activity, which 50% of the population survived 10

11 (WA 50 ). For non-adapted populations, approximately DJs were subjected to different a w -values (ranges between 0.99 to 0.8). After 24h of incubation at 25 C in the dark, DJs were washed from PEG with Ringer s solution on 10 μm mesh size sieves by vacuum suction equipment and then transferred to storage solution (11.25 g NaCl, g CaCl 2 2H 2 O, g MgSO 4, 0.1 g ascorbic acid, ad 1 l aqua dest.) for rehydration and incubated for another 24h. The number of surviving and dead nematodes was counted with an inverted light microscope. For the adaptation phase, DJs were pre-adapted at a w -value of 0.95 for 48h and then transferred to different concentrations of PEG to achieve the a w -values ranging between 0.9 to 0.52 and DJs were incubated for 24h in these solutions. They were kept in storage solution for rehydration for 24h and then survival was recorded. The assessment of desiccation tolerance was repeated at least three times with different batches. ESTABLISHMENT OF MONOXENIC LIQUID CULTURE In order to produce monoxenic in vitro cultures, bacteria-free J1 nematodes were combined with symbiotic bacteria according to Lunau et al. (1993). Egg isolation was done with the original strain HYB13. DJs were inoculated to Galleria larvae ( DJs/ larva), kept at 25 C. Infected larvae were teared apart between 3-5 days post inoculation to check for the presence of gravid females. After 5 days, first gravid females were seen and collected (200 individuals) into staining block, washed several times with sterile Ringer s solution in order to reduce bacterial contamination. They were placed in glass tubes together with razor blade pieces and vortexed until the eggs were released. The eggs were collected passing the liquid through a 50 µm sieve, the vial was centrifuged (2000 rpm, 2 min) with Ringer s solution for the replacement of supernatant. Surface sterilisation of the eggs was done using sterilisation solution (0.5 ml Sodiumhypochloride NaOCl 12%, 1.5 ml 4 mol. NaOH, 10 ml aqua dest). After shaking 4 min in sterilisation solution, eggs were centrifuged and washed with YS medium. The YS broth with the eggs was transferred into sterile 24 multiwell plates and kept 11

12 at room temperature for 48h. Then, freshly hatched bacteria free J1s were transferred into Wouts agar medium (16.0 g Bacto Nutrient Broth, 12.0 g Bacto Agar, 5.0 g sun-flower oil, ad 1 l aqu dest.) seeded with the symbiotic bacteria X. bovienii and kept for two weeks. After that time, the population development of the nematodes was completed and only DJs remained. Then, they were scaled up to 200 ml flasks into a nematode growing medium (per liter distilled water: 15 g yeast extract, 15 g soya flour, 4 g NaCl, 0.35 g KCl, 0.3 g CaCl 2, 0.2 g MgSO 4, 30 g sunflower oil) which was pre-inoculated with the symbiotic bacterium X. bovienii. Culture flasks were kept on a rotary shaker at 25 C for two weeks then stored at 4 C in the cooling room. ESTABLISHMENT OF INBRED LINES Propagation of single kept gravid females on Wouts agar Strains HYB13 and Com were used for this experiment. For inbreeding, single females were kept on Wouts agar in Petri dishes (60 x 55 mm diameter). The dishes were inoculated with the symbiotic bacteria X. bovienii. Bacteria culture was obtained from e-nema GmbH and stored with 15% sterilized glycerol at -80 C in Eppendorf tubes. Before using for the experiment, they were incubated into YS medium (5.0 g yeast extract, 5.0 g NaCl, 0.5 g NH 4 H 2 PO 4, 0.5 g K 2 HPO 4, 0.2 MgSO 4 7H 2 O, ad 1l aqua dest.) for 24h and kept on rotary shaker at 25 C in dark room. Inbreeding regimes Four different inbreeding regimes have been used in this study. In the first and the second the nematodes were obtained from in vivo culture. In the second regime, the single kept females were treated with antibiotics before they were inoculated to the Wouts agar plates in order to reduce microbial contaminations. In the third regime nematodes from monoxenic liquid culture were used. This was also done in order to reduce microbial 12

13 contaminations with non-symbiotic bacteria. In the 4 th regime the nematodes had been selected for desiccation tolerance before inbreeding. These nematodes had been taken again from in vivo cultures because of limited time for the establishment of in vitro cultures. For each regime, 50 inbred lines were produced. All experiments were carried out at 25 C. The agar plates from every second inbreeding step were kept at 25 C until new DJs developed. After that the nematodes were kept at 4 C until further use. Ten lines were chosen randomly and these nematodes were later used for the infection of Galleria larvae in order to produce fresh DJs for the assessment of the emergence rate from infected insects, the virulence and the desiccation tolerance. 1 st regime: Inbreeding started from in vivo culture without antibiotic treatment 1. Infection of Galleria larvae with the basic nematode strain (HYB13, Com) with DJs per larva 2. After 5 days, single gravid females were picked from dissected Galleria larvae, washed in sterile Ringer s solution and transferred to Wouts agar plates inoculated with the symbiotic bacteria the same day 3. After another 5 days, single gravid females were picked from the Wouts agar and transferred directly to new agar plates inoculated with symbiotic bacteria. 4. This was carried out up to the fourth inbreeding step. After the second inbreeding step the Wouts agar plates had been inoculated one day before nematode transfer 2 nd regime: Inbreeding start from in vivo culture with antibiotic treatment of nematodes 1. Due to contamination of the nematode cultures causing loss of inbred lines, nematodes were treated with antibiotics prior to transfer to axenic symbiont bacterial cultures. Single gravid females (HYB13, Com) from the 4 th inbreeding step of the first 13

14 inbreeding regime were exposed to antibiotic solution (streptomycin 100 µg/ml, ampicillin 300 µg/ml, chloromophil 60 µg/ml, kanamycin 100 µg/ml, erythromycin 200 µg/ml mixed in 50 ml YS medium) 2. Individual females were kept in antibiotic solution (200 µl) using 24 multiwell plate for 24h in the dark 3. After 24h incubation, they were washed in sterile Ringer s solution and transferred to new agar plates pre inoculated with symbiotic bacteria 4. After five days, single gravid females were transferred to new agar plates again 5. This was carried out up to the 8 th inbreeding step 3 rd regime: Inbreeding start from monoxenic in vitro culture DJs/ Petri dish, taken from liquid culture, were incubated on agar plates preincubated with the bacterium X. bovienii 2. After five days, gravid females were picked and transferred singly to a new agar plates 3. The same process was applied up to the 8 th inbreeding step 4 th regime: Inbreeding start from in vivo culture including a selection step 1. Ten thousand DJs from the basic strain of HYB13 were exposed to desiccation stress at an a w -value of 0.88 at which 10% of the populations can survive 2. After 24h, they were rehydrated in storage solution and left overnight 3. Then the surviving DJs were taken for the infection of Galleria larvae 4. After 5 days, gravid females were picked from dissected larvae, washed with sterile Ringer s solution and singly transferred to agar plates pre-incubated with the symbiotic bacterium X. bovienii 14

15 5. Transfer of new generation gravid females was done again after 5 days to a new agar plate 6. Transfer continued up to the 2 nd inbreeding step EMERGENCE TEST The DJs developed from each inbred line (single kept female) were used to infect Galleria larvae. The percentage of infected Galleria larvae was determined from which finally new DJs emerged. From each inbred line, 8 Galleria larvae have been infected with DJs per insect. VIRULENCE TEST HYB13 and Com strain nematodes were used for an infectivity assessment in order to observe differences in virulence before and after inbreeding. Inbred lines of the fourth inbreeding step were taken for comparison with the initial strain. Before the virulence test could be started, the DJs resulting from the 4 th inbreeding step needed to propagate in G. mellonella. The emerging DJs were used for the test. This was done since the DJs stored on agar plates usually have a lower quality than those freshly emerged from larvae (pers. communication Olaf Strauch). Virulence was tested using the last instars larvae of lesser mealworm (Tenebrio molitor L.) (Coleoptera: Tenebrionidae). Two hundred gram of sand was mixed with 10% of tap water (180 g sand + 20 g soil) and placed in 150 x 20 mm size of plastic Petri dish. Twenty mealworms were added per dish. Different dosages of DJs (5, 10, 20, 50, 80, and 100 in 1 ml tap water per plate) were used to estimate the LD 50 (dosage at which 50% of the insect larvae were killed). Control treatment was included with 1 ml tap water. The Petri dishes were kept at 25 C in the dark. Insect mortality was checked after one week of incubation. 15

16 DATA ANALYSES Desiccation tolerance To calculate the mean desiccation tolerance within the strains, original data (percentage of survival DJs versus water activity) were fitted to a cumulative normal distribution (Fig. 1). The fitting was done by minimizing the chi-square through comparing the original data and the theoretical normal distribution. The mean and the standard deviation of the fitted normal distribution was used to determine the mean tolerated a w -value (WA 50 ) and the a w -value at which the most desiccation tolerant 10% of individuals survived (WA 10 ). Comparison of strains was done based on the WA 50 and the population of WA 10 was used for selection. Analyses of variance (ANOVA) and the Tukey s HSD test were done to evaluate the significant differences between the strains. 16

17 Survival (%) S. Türköz, % 75% original data 50% 25% Cumulative normal distribution WA10 WA50 0% 0,7 0,75 0,8 0,85 0,9 0,95 1 a w -value Fig. 1. Estimation of the desiccation tolerance of entomopathogenic nematodes by fitting mortality data to a cumulative normal distribution. The data for the given example were obtained from a non-adapted population of S. riobrave strain Sr Lin4. Percentage of survival after exposure to different water activities (a w -value) and a fitted cumulative normal distribution were used to calculate the mean tolerated water activity WA 50 and the a w -value tolerated by the most tolerant 10% of the nematode population (WA 10 ). Virulence test In order to calculate the LD 50 (population at which 50% survive) original data were fitted to a saturation curve (Fig. 2) by minimizing the chi-square for the comparison with the theoretical distribution. Abbot corrected mortality was used for different dosages. 17

18 Lethal concentration (%) S. Türköz, ,75 0,5 0, DJs/ Mealworm Fig. 2. Estimation of LD 50 which is required to kill 50% of the population by fitting the original data to saturation curve. The data for the given example were obtained from the basic strain of HYB13. Dashed lines indicate the LD 50. Results COMPARISION OF THE DESICCATION TOLERANCE OF SEVERAL STEINERNEMA SPECIES AND STRAINS Three S. riobrave strains, 2 S. feltiae strains and 1 S. yirgalemense strain was tested and compared regarding their desiccation tolerance. There were significant differences found between the adapted and non-adapted populations, adapted populations representing a higher tolerance compared to the non-adapted nematodes both for mean tolerated water activity and the best 10% individuals (ANOVA, WA 50 : F = , df = 6, 49, p < ; WA 10 : F= , df = 6, 49, p < ) (Fig. 3 and 4). For the mean tolerated a w -value of non-adapted populations, there was no significant differences between the species (Tukey s HSD test, df = 27, p 0.05) (Fig. 3). After adaptation, the WA 50 was significantly lower for Sy 157-c and Sr Lin 4 compared to the commercial Sf strain but no significant difference was recorded 18

19 compared to the HYB strain (Tukey s HSD test, df = 27, p 0.05). Strain Sr Lin4 was the most tolerant to desiccation (a w -value = 0.73) among the all strain. The Com strain was the least tolerant (a w -value= 0.912). Fig. 3. Mean tolerated water activity (WA 50 ) of adapted and non-adapted populations of S. feltiae (Sf), S. riobrave (Sr) and S. yirgalemense (Sy) strains. Error bars on the column represent standard deviation of different replicates. Columns with the same latter are not significantly different (Tukey s HSD test, p 0.05). Most desiccation tolerant individuals were found in Sr 3-7 strain (a w -value= 0.52) with an adaptation phase (Fig. 4). The WA 10 value for this strain was significantly lower than that recorded for Sf HYB and Sy 157-c (Tukey s HSD test, df = 27, p 0.05). 19

20 Fig. 4. Water activity tolerated by only 10% of the population (WA 10 ) of adapted and nonadapted populations of S. feltiae (Sf), S. riobrave (Sr) and S. yirgalemense (Sy) strains. Error bars on the column represent standard deviation of different replicates. Columns with the same latter are not significantly different (Tukey s HSD test, p 0.05). INBREEDING Emergence of inbred lines The percentage of infected Galleria larvae was determined from which finally new DJs emerged. Different treatments were compared in terms of emergence. Antibiotic application was done from 4 th inbreeding step onwards, both for Com and HYB strain, thus no data are available with antibiotic application for individuals originating from the 2 nd inbreeding step. There was no emergence from Galleria infected with DJs from the antibiotic application after the 8 th inbreeding step for the Com strain. 20

21 In general, the number of emergence from inbred lines was low for each treatment (Fig. 5 and 6). The emergence of the Com strain was significantly higher than of the HYB strain (Tukey s HSD test, df= 82, p= 0.032). For the Com strain, the emergence without antibiotic application on the second inbreeding step (35 %) was significantly higher compared to those which had been treated with antibiotics on the 4 th and 8 th inbreeding step (Tukey s HSD test, df= 76, p 0.05). In general, there was a gradual decline in emergence with further inbreeding steps for the lines that were not treated with antibiotics although it was not significant (Fig. 5). Fig. 5. Percentage of emergence (%) from dead insects on each inbreeding step of the Com strain. Columns with the same letter are not significantly different compared to the other treatments (Tukey s HSD test, p 0.05). 21

22 For the HYB strain, no significant differences were recorded between the different treatments (Tukey s HSD test, df= 76, p 0.05) (Fig. 6). There was no emergence obtained from insects infected with nematodes resulting from the 8 th inbreeding step. The highest emergence (20%) was obtained from insects infected with nematodes that had been treated with antibiotics from the fourth inbreeding step. The lowest emergence was obtained from nematodes originating from monoxenic liquid culture. Fig. 6. Percentage of emergence (%) from dead insects on each inbreeding step of the HYB strain. Without antibiotic application (-, +) refers to inbred lines without or with prior selection to desiccation stress. Columns with the same letter are not significantly different compared to the other treatments (Tukey s HSD test, p 0.05). 22

23 Survival of inbred lines The maximal obtainable inbreeding step for each inbreeding line was assessed before the line died out. Antibiotic application could not significantly influence survival of inbred lines (Fig. 7). However, significant differences were recorded between the strains (ANOVA, df= 3, 194; F= ; p < ), Com showed a significant better survival than the HYB13 whether treated with antibiotic or not (Tukey s HSD test, df= 194; p < ). Using nematodes from monoxenic liquid culture for the establishment of inbred lines was significantly more successful than with nematodes from in vivo cultures (Tukey s HSD test, df= 146; p < ). Fig. 7. Maximum inbreeding steps reached after different treatments for the HYB13 and Com strains. Error bars on the columns represent standard deviation of different replicates. Columns with the same letter are not significantly different from other treatments (Tukey s HSD test, p 0.05) 23

24 Desiccation tolerance of inbred lines Tolerance to desiccation was summarized for inbred lines after different inbreeding step and compared with the original strain. Because of the low emergence rate from Galleria larvae infected with inbred lines it was not possible, in most cases, to carry out at least 3 replicates for the desiccation tolerance assessments. For the Com strain, statistical comparisons could be only done for the initial strain and the lines 5, 7, 13, 16, 20, 42, 48, and 49. Mean tolerated water activity (WA 50 ) of non-adapted populations did not differ significantly from the original strain and between the lines (WA 50 : F = 1.584, df = 8, 21, p = 0.189) (Fig. 8). When comparing the different inbreeding steps, in most cases desiccation tolerance remained the same up to six inbreeding steps for different lines. There was a change observed only on the eighth inbreeding step although it was not significant. For adapted populations, there was already some increase obtained at the second inbreeding step and it was more pronounced when moving to the sixth inbreeding step (Fig. 9). Line 7, after six inbreeding steps, was significantly higher in desiccation tolerance than the original strain and the line 49 from the same inbreeding step (WA 50 : F = 3.459, df = 8, 21, p 0.05). In general, the best 10% individuals (WA 10 ) tolerated lower water activity values than the average populations. For non-adapted populations, inbred lines were less tolerant than the initial strain but no significant difference was found (WA 10 : F = 2.287, df = 8, 21, p = 0.062) (Fig. 10). For the adapted populations, there was no significant difference found to the initial line. Significant differences were found between the line, line 7 after six inbreeding step was significantly more tolerant than line 49 on the same inbreeding step and line 5 on fourth inbreeding step (WA 10 : F = 3.055, df = 8, 21, p = 0.019) (Fig. 11). Some of the lines 24

25 decreased in desiccation tolerance even after two inbreeding step although it was not significant. Fig. 8. Mean tolerated water activity (WA 50 ) of non-adapted populations of S. feltiae (strain Com). Error bars on the columns represent standard deviation of different replicates for the lines 0, 5, 7, 13, 16, 20, 42, 48, and 49. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. 25

26 Fig. 9. Mean tolerated water activity (WA 50 ) of adapted populations of S. feltiae (strain Com). Error bars on the column represent standard deviation of different replicates for the lines 0, 5, 7, 13, 16, 20, 42, 48, and 49. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain.. 26

27 Fig. 10. Water activity tolerated by only 10% of the population (WA 10 ) of non-adapted population of S. feltiae (strain Com). Error bars on the column represent standard deviation of different replicates for the lines 0, 5, 7, 13, 16, 20, 42, 48, and 49. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. 27

28 Fig. 11. Water activity tolerated by only 10% of the population (WA 10 ) of adapted population of S. feltiae (strain Com). Error bars on the column represent standard deviation of different replicates for the lines 0, 5, 7, 13, 16, 20, 42, 48, and 49. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. For the HYB strain population without selection pressure analyses of variance (ANOVA) test could be done for only the initial strain and line numbers 17 and 27 from the second inbreeding step. No significant differences could be obtained between these lines and the initial strain. Both the mean tolerance (WA 50 ) (ranging between 0.7 to 0.899) and the most tolerant 10% individuals (WA 10 ) (ranging between and 0.804) were lower when the DJs had gone through an adaptation phase (Fig. 13 and 15). When comparing all lines from different inbreeding steps, there was already an increase obtained for the mean desiccation 28

29 tolerance (WA 50 ) of pre-adapted nematodes after the second inbreeding step and it further improved although differences were not significant. Great variation was observed from one line to another for those, which were in the same inbreeding step for WA 10 of adapted nematodes (Fig. 15). Fig. 12. Mean tolerated water activity (WA 50 ) of non-adapted populations of S. feltiae (strain HYB) without selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 17 and 27. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. 29

30 Fig. 13. Mean tolerated water activity (WA 50 ) of adapted populations of S. feltiae (strain HYB) without selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 17 and 27. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. 30

31 Fig. 14. Water activity tolerated by only 10% of the population (WA 10 ) of non-adapted populations of S. feltiae (strain HYB) without selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 17 and 27. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain.. 31

32 Fig. 15. Water activity tolerated by only 10% of the population (WA 10 ) of adapted populations of S. feltiae (strain HYB) without selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 17 and 27. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. For the inbred lines of the HYB strain, which had first been exposed to desiccation stress (with selection pressure), line numbers 1, 14, 27 and 38 could be used for statistical comparison with the initial strain. Desiccation tolerance was assessed after two inbreeding step for the selected inbred lines. For non-adapted population, significant differences were recorded (WA 50 : F = 5.389, df = 4, 12, p 0.05) for the WA 50 for the line number 1 and 38 when compared with the original population (Fig. 16). Significant differences were observed between the lines (line 27 to 1 and 38) (Tukey s HSD test, df= 12, p 0.05). No significant 32

33 effect was observed for desiccation tolerance between the different lines of pre-adapted populations (WA 50 : F = 0.126, df = 4, 12, p = 0.126) (Fig. 17). Fig. 16. Mean tolerated water activity (WA 50 ) of non-adapted populations of S. feltiae (strain HYB) with selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 2, 14, 27 and 38. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. 33

34 Fig. 17. Mean tolerated water activity (WA 50 ) of adapted populations of S. feltiae (strain HYB) with selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 2, 14, 27 and 38. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. In general, the best 10% individuals (WA 10 ) tolerated lower water activity values than the average populations. For the non-adapted population significant differences were recorded for the line numbers 1 and 38 compared to the initial strain (WA 10 : F = 5.496, df = 4, 12, p 0.05) (Fig. 18). The highest desiccation tolerance was obtained with line 38 (a w = 0.734) for non-adapted DJs. Significant differences were observed between line 1 and 27 (Tukey s HSD test, df= 12, p 0.05). With inclusion of an adaptation phase better tolerance to desiccation was achieved, but it was not significant (Tukey s HSD test, df= 12 p 0.05) (Fig. 19). 34

35 Fig. 18. Water activity tolerated by only 10% of the population (WA 10 ) of non-adapted population of S. feltiae (strain HYB) with selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 2, 14, 27 and 38. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. 35

36 Fig. 19. Water activity tolerated by only 10% of the population (WA 10 ) of adapted population of S. feltiae (strain HYB) with selection pressure. Error bars on the column represent standard deviation of different replicates for the lines 0, 2, 14, 27 and 38. Columns with the same letter are not significantly different (Tukey s HSD test, p 0.05). The dashed line is the reference point recorded for the initial strain. HYB and the Com strain were compared using analyses of variance (ANOVA). Significant difference was observed for the mean tolerated water activity of non-adapted populations, HYB showing better tolerance than the Com strain (WA 50 : F = 2.267, df = 8, 63, p 0.05). In order to see whether there is a difference with the inclusion of selection phase, a two sample t test comparison was done at 95% confidence interval for the HYB strain. Significant differences were observed between treatments without and with selection pressure for WA 10 for both, adapted and non-adapted populations (two sample t- test: df = 30, t= 2.042, p= 0.037). 36

37 Whether a correlation between the desiccation tolerance and the inbreeding steps can be detected was evaluated. Not all the lines could be assessed due to unsuccessful propagation in the insects. Lines that could be used for further inbreeding step were taken into consideration, in case of the HYB strain the lines 16, 17 and 44 and for strain Com lines 5 and 20. In general, there is a correlation found between the number of inbreeding steps and the desiccation tolerance. Most of the lines followed the trend of increasing desiccation tolerance with increasing number of inbreeding steps. Pearson correlation and regression of variance was determined and compared at a significance level of p Mean tolerated water activity was observed for adapted and non-adapted populations of both strains. For the adapted population, HYB strain showed an enhanced desiccation tolerance for all the three lines up to fourth inbreeding step and a loss of tolerance of line 44 after the eighth inbreeding step (Fig. 20). There was a significant correlation found for the lines 16 (Pearson correlation test: p= 0.027) and 17 (Pearson correlation test: p= 0.018), jumping from 0.83 to and 0.658, respectively. Non-adapted populations followed the same trend (Fig. 21). No significant differences were seen for the non-adapted nematodes. 37

38 a w values a w values S. Türköz, ,95 HYB WA 50 Adapted 0,9 0,85 0,8 0,75 0,7 0,65 16 (r = *) 17 (r = *) 44 (r = 0.53) 0, Inbreeding steps Fig. 20. Correlation between the mean tolerated water activity (WA 50 ) and the inbreeding step of the HYB strain of S. feltiae for adapted populations for lines 16, 17 and 44. R is the coefficient of variance from the regression analysis and * refers for significance. 0,95 HYB WA 50 Non-adapted 0,9 0,85 0,8 0,75 0,7 16 (r = ) 17 (r= ) 44 (r = ) 0,65 0, Inbreeding steps Fig. 21. Correlation between the mean tolerated water activity (WA 50 ) and the inbreeding step of the HYB strain of S. feltiae for non-adapted populations for lines 16, 17 and 44. R is the coefficient of variance from the regression analysis. 38

39 a w values S. Türköz, 2013 For strain Com the same trend with decreasing a w -values after continuous inbreeding was recorded for adapted and non-adapted populations (Fig. 22 and 23). The regression coefficient was negative indicating an increase in the desiccation tolerance. The mean tolerated a w -value of adapted populations for line 20 dropped from 0.91 to 0.68 significantly (Pearson correlation test: p= 0.006). No significant differences were assessed for non-adapted populations (Pearson correlation test: p 0.05). 0,95 Com WA 50 Adapted 0,9 0,85 0,8 0,75 0,7 5 (r = ) 20 (r = *) 0,65 0, Inbreeding steps Fig. 22. Correlation between the mean tolerated water activity (WA 50 ) and the inbreeding step of the Com strain of S. feltiae for adapted populations for lines 5 and 20. R is the coefficient of variance from the regression analysis and * refers for significance. 39

40 a w values S. Türköz, ,95 Com WA50 Non- adapted 0,9 0,85 0,8 0,75 0,7 5 (r = ) 20 (r = ) 0,65 0, Inbreeding steps Fig. 23. Correlation between the mean tolerated water activity (WA 50 ) and the inbreeding step of the Com strain of S. feltiae for non-adapted populations for lines 5 and 20. R is the coefficient of variance from the regression analysis. Correlation between the mean variability and the inbreeding steps was checked as well. Correlation was found for the mean tolerated water activity of adapted populations for strain Com. In most cases, there was no correlation recorded due to the great variance between the replicates. Results on Infectivity The infectivity of Com and HYB strains were determined after the 4 th inbreeding step and compared with the initial strain. The HYB strain with inclusion of selection pressure was also included in order to see the differences in relation to non-selected lines. Analyses of variance could not be assessed due to the missing replicates for both strains. For the Com strain only the line numbers 4, 5, 13, 15 and 16, for the HYB strain without selection pressure 40

41 only two lines 19 and 23, for the HYB with selection pressure lines 1, 14, 27 and 38 were assessed. T- test comparison was only done for the initial strain and the line 13 of the Com strain, for which three replicates could be assessed. In most cases infectivity of the inbred lines increased after the 4 th inbreeding step for the Com strain. Line 13 was significantly higher in virulence compared to the initial strain (two sample t- test : df= 5, t= 2.571, p < 0.008) (Fig. 24). Fig. 24. Lethal dose at which 50% (LD 50 ) of Tenebrio mollitor larvae were killed by S. feltiae strain Com. Different letter represent significant differences (two sample t- test : p 0.05). For the HYB strain infectivity seems to be reduced whether with selection pressure or not. No correlation was found between the infectivity and the desiccation tolerance (Fig. 25). 41

42 HYB without selection pressure LD HYB Line numbers HYB with selection pressure LD HYB Line numbers Fig. 25. Results for HYB strain with and without selection pressure on Lethal dose at which 50% (LD 50 ) of Tenebrio mollitor larvae were killed. Discussion Entomopathogenic nematodes are widely used in biological control against pest insects in cryptic environments (Grewal et al., 2005). The commercial exploitation of these 42

43 nematodes motivated research and development targeting at genetic improvement of beneficial traits (Anbesse et al., 2013b). Progress in genetic improvement of beneficial traits of entomopathogenic nematodes (Steinernema and Heterorhabditis) by hybridisation and genetic selection (e.g., Mukuka et al., 2010; Nimkingrat et al., 2013a,b) has often been lost again once the selection pressure was released during mass production (Anbesse et al., 2012a,b; Nimkingrat et al., 2013a, b). Recently, a method was developed for Heterorhabditis bacteriophora to stabilize traits by selection of improved inbred lines which maintain enhancements reached by genetic selection (Anbesse et al., 2012a, b). Culturing this nematode in liquid media results in highly homozygous inbred lines. Under liquid culture conditions the nematode reproduces exclusively through self-fertilisation of hermaphrodites (Strauch et al., 1994; Johnigk et al., 2002). This approach is not possible for Steinernema spp. because they are amphimictic. Therefore, objective of this study was, to first of all develop a procedure to produce inbred lines of S. feltiae. Inbreeding in Steinernematid nematodes can only be done by crossing offspring of one female and continuously cross brother and sisters of one parent. To achieve this goal, single individuals had to be cultured separately. This can only be done in in vitro culture. Dauer juveniles were inoculated into G. mellonella and when they had developed to adults, females were transferred to agar plates seeded with the symbiotic bacterium X. bovienii. This approach was often not successful and many lines were lost after a few cycles because of contamination of the in vitro cultures. The contaminants were introduced from the insect cadavers. To overcome this problem, females were treated with antibiotics before transfer into the X. bovienii cultures. However, also this approach often failed as the treatment could not completely eliminate the contaminants. Finally, monoxenic cultures of S. feltiae were produced and the whole procedure was done under in vitro conditions. This approach was more successful and is therefore recommended for future production of inbred lines of 43