Isolation and Identification of Kairomone(s) in the Daphnia-Scenedesmus System

Infochemicals play an important role in interactions between living organisms in aquatic environments. Although the presence of these chemical cues is confirmed in more and more systems, the chemical structures of the compounds involved remain predominantly elusive and the identification of these compounds is essential to advance the research on chemical communication. An overview of chemical cues involving Daphnia (either as producer or receiver) is given and the progress towards their isolation and structure elucidation is described (Chapter 1). Most of the research so far has concentrated on the elucidation of kairomones produced by predators of Daphnia (especially Chaoborus and several species of fish). Less study has been devoted to the isolation of the infochemical exuded by Daphnia that causes colony formation in its prey Scenedesmus. One of the main aims of this study was the isolation and identification of this chemical cue. Colony formation in Scenedesmus only occurs when unicellular populations are exposed to either Daphnia or water that had contained Daphnia. It was concluded that the responsible cue had a chemical rather than a mechanical nature, since filtered Daphnia water also showed the colony formation activity. This colony formation was the basis for the development of a bioassay (Chapter 2). A bioassay is a test that is used to measure biological activity (in this case colony formation) of chemical mixtures or biological parameters. Colonies are indicated by high values and single cells are indicated by low values. Unfortunately over time a gradual decline of the difference between negative and positive controls was observed and efforts were undertaken to determine the cause for this decline. Several conditions were investigated (such as time, temperature, algae strain, culture medium, location, incubator, Erlenmeyer size, bacterial growth and microevolution). Additionally some general properties of the kairomone (such as thermal decomposition, biodegradation and concentration) were tested. A correlation between any of the above mentioned factors and the gradual decline of the difference between negative and positive controls was not found. Given that the bioassay was performed under such highly variable and not strictly controlled circumstances, this particular bioassay seems to be rather robust. However this does not defer from the fact that the quality of the bioassay did decline over time. Until the variable is identified that is responsible for the observed decline in difference between positive and negative controls, more care should be taken to standardise as many variables as possible. Despite its drawbacks, a bioassay still remains the best option to guide isolations of bioactive compounds through controlled experiments as long as observed differences are statistically significant. To find the most suitable and practical method for the analysis of Daphnia test water several sample pre-treatment methods were compared (Chapter 3), such as liquid-liquid extraction, solid-phase extraction, stir-bar sorptive extraction, solid-phase disk extraction. A test mixture with ten known natural compounds differing in polarity (log K _o/w between -4.34 and 3.70) was used. The best method for small amounts of sample was either 'stirred' SPE or 'cartridge' SPE, but for large amounts of sample (sometimes up to 20 L) 'syringe' SPE was more suited. Consequently an SPE analysis protocol was developed that could elute the active compound in one fraction (Chapter 4). The experiments were performed with different concentrations of organic solvents and different sorbents (endcapped C ₁₈ , MF C ₁₈ , non-endcapped C ₁₈ , C ₈ , C _2, CN, ENV ⁺ and Oasis ^® HLB). Endcapped C ₁₈ was eventually chosen for further experiments (other sorbents did not perform better) and extracted with differing concentrations of methanol in water (50%, 85%) and pure methanol (100%). The chemical cue was most often recovered from the 85% aqueous methanol fraction, which indicates the cue is moderately non-polar.Biological activity was lost when active Daphnia water was partitioned at pH 12.0 against ethyl acetate. The aqueous and organic layer were both inactive, either by inactivation of the kairomone by the basic conditions in the aqueous layer or possibly more than one compound is present with synergistic effects. At lower pH (2.0 and 7.0) biological activity was recovered from the organic layer. This could be an indication that the active compound contains an anionic group. Experiments performed with ion exchange materials (SAX, SCX and Amberlite IRA-400) focused initially on the anion exchanger (SAX). However colony formation activity was recovered from the unretained fraction in contrast to what had been reported previously. This unexpected result prompted extraction with a cation exchanger (SCX). To exclude problems related to pH sensitive silica based sorbents, experiments were repeated on a resin based sorbent (Amberlite IRA-400),however a similar result was obtained as with the SAX sorbent. No satisfying explanation was found for the presence of biological activity in the unretained fractions and absence from the retained fractions, but different counterions on the ion exchangers could play a role. The enriched extracts obtained by SPE (C ₁₈ ) were fractionated by high performance liquid chromatography. One fraction showed a significantly higher biological activity relative to the control ('Fraction C'). Further fractionation yielded three active fractions (C2, C3,C6). This could be an indication that more than one compound is responsible for the activity. Several natural products were biologically inactive when screened in the bioassay. They were, ecdysterone and juvenile hormone III (important hormones in other Crustaceae ), urea (proposed as kairomone in Daphnia - Scenedesmus system ), and geranic acid (reference compound). Although an assumption was made to ignore possible synergistic or additive effects in these experiments, the possibility of synergism or additivity should not be ignored, given that most likely more than one active fraction is present. At this point, due to the lack of reproducible and significant results from bioassay-guided separations, another way to identify possible candidates for the role of kairomone had to be used. Daphnia and control water were first extracted using SPE and then analysed with chromatographic techniques. Chromatograms of biologically active extracts were then compared with chromatograms of non-active control extracts to determine and recognise unique peaks (i.e. peaks only present in active Daphnia test water extracts). In an attempt to maximise the available data on the unknown colony inducing compound(s) several techniques were applied simultaneously, such as gas and liquid chromatography (Chapter 5). Several small unique peaks were recognised in the silylated extracts of Daphnia test water with GC-MS analysis. Some of these were tentatively identified as dodecanol, azelaic acid, sebacic acid and veratroylformic acid, but they did not induce colonisation. HPLC detection was performed not only with ultraviolet spectroscopy but also with evaporative light scattering and by electrospray ionisation-mass spectrometry to avoid overlooking compounds without a UV chromophore. LC analysis on four columns with different packings ensured that peaks were well separated on at least one column. Chromatograms with the best resolution were obtained on a C ₁₈ column with an ACN-H ₂ Ogradient and UV detection. High noise levels reduced the usefulness of ELS detection. Several peaks unique to 90% aqueous MeOH extracts of Daphnia test water were detected, but unfortunately not identified. Some of the recognised unique peaks ( B, G,K ) eluted in previously identified active regions ('Fraction C'). Especially peak B ([M-H] ^¯ = 752.8 ?,l_max227 nm) was present in high amounts and well separated from neighbouring peaks. Therefore this peak was further analysed by liquid chromatography-nuclear magnetic resonance. Peaks from several extracts were trapped onto one SPE cartridge. This way a sufficient amount of analyte could be transferred into the NMR probe to allow recording of a 1-dimensional ¹ H-spectrum. Unfortunately the spectrum did not lead to elucidation of the structure of peak B . One aliquot of this collected fraction was therefore analysed by high-resolution mass spectrometry and liquid-chromatography-quadrupole time-of-flight mass-spectrometry to obtain an accurate mass, while another aliquot was checked for biological activity in a bioassay. Unfortunately analysis with HRMS was unsuccessful and analysis with LC-QTOF has not yet yielded results. The peak with a possible pseudo molecular mass of 752.8 ([M-H] ^¯) could not be detected. The other aliquot that was tested for biological activity in the bioassay showed significant differences between the negative control, positive control and peak B . This peak could therefore play a role in the induction of colonies in Scenedesmus , although it is still unclear whether it acts alone. Should peak B prove to be (partly) responsible for colony formation in Scenedesmus then the most important objective of this study has been partly reached, namely the isolation of kairomone(s) in the Daphnia - Scenedesmus system. This information will enable and facilitate research into the other objectives.