Effects of nanomolar cadmium concentrations on water plants – comparison of biochemical and biophysical mechanisms of toxicity
under environmentally relevant conditions
Dissertation zur Erlangung des
Akademischen Grades eines Doktors der Naturwissenschaften
vorgelegt von Andresen, Elisa
an der
Fachbereich Biologie
Mathematisch-Naturwissenschaftliche Sektion
Tag der mündlichen Prüfung: 21.02.2014 1. Referent: Prof. Dr. Hendrik Küpper
2. Referent: Prof. Dr. Eva Freisinger
Table of contents
Summary ... 4
Zusammenfassung ... 5
Abbreviations ... 8
1. Introduction ... 9
1.1. Heavy metals in plants ... 9
1.2. Cadmium ... 9
1.3. Environmentally relevant experiments ... 15
1.4. Setting of the experiments in this thesis ... 16
1.5. Fluorescence kinetic measurements ... 17
1.6. Aims of this study ... 20
2. Publications in peer-reviewed journals and manuscripts ... 21
2.1. Cadmium toxicity investigated on physiological and biophysical level under environmentally relevant conditions using the aquatic model plant Ceratophyllum demersum L. ... 21
2.2. Different strategies of cadmium detoxification in the submerged macrophyte Ceratophyllum demersum L. ... 49
2.3. Effects of Cd & Ni toxicity to Ceratophyllum demersum under environmentally relevant conditions in soft & hard water including a German lake ... 73
3. General discussion, conclusions and perspectives for future work ... 106
3.1. Cadmium toxicity ... 106
3.2. Cadmium and zinc – a complex relationship ... 110
3.3. Cadmium induces reactive oxygen species ... 111
3.4. Cadmium detoxification ... 112
4. References ... 115
5. Appendix ... 136
5.1. Author contributions ... 136
5.2. Acknowledgments ... 138
Summary
In this thesis, the effects of the highly toxic heavy metal cadmium (Cd) on the rootless aquatic model plant Ceratophyllum demersum are investigated on the biochemical and biophysical level. The experiments were carried out using environmentally relevant conditions, i.e. light and temperature followed a sinusoidal cycle, a low biomass to water ratio resembled the situation in oligotrophic lakes and a continuous exchange of the defined nutrient solution ensured that metal uptake into the plant was not limited by the nutrient solution, but the capacity of the plant. Above all, Cd concentrations in the nanomolar range were applied and experiments lasted long enough to observe chronic toxicity.
The toxicity study revealed that the first site of inhibition was the photosynthetic apparatus. The maximal quantum efficiency of photosystem II (PS II) photochemistry in dark adapted state as well as the PS II operating efficiency in actinic light were the first parameters to be reduced. Only afterwards, an increase in reactive oxygen species (ROS) was observed, indicating that they are the result and not the cause of dysfunctional photosynthesis. For most affected parameters, the respective threshold concentration of inhibition or upregulation due to Cd treatment was 20 nM. This is a much lower concentration than applied in many previous studies. All of the observed effects were more pronounced in plants subjected to Cd stress under high light conditions compared to low light conditions, suggesting a protective role of the comparatively larger antenna system of low light grown plants.
Cadmium treatment led to a redistribution of other metals, especially Zn in the tissue of C. demersum as revealed using the non-invasive technique of micro X-ray fluorescence (µ-XRF) on frozen hydrated leaves. At low Cd concentrations, Zn was found mainly in the epidermis and in the mesophyll. At moderately toxic Cd concentration (20 nM), a higher proportion of Zn was found in the mesophyll. At the highest Cd concentrations, Zn was seemingly stuck in the vein, suggesting that the Zn-exporters were blocked by Cd.
As part of the detoxification process, a changing distribution of Cd with increasing Cd concentrations was observed. No Cd was detected in the plants from the control treatment (no Cd added) and a homogenous distribution of Cd all over the leaf section was revealed at low Cd treatment. However, at moderately toxic and highly toxic Cd concentrations, sequestration of Cd into specific organs and tissues was observed. The process of sequestration, transport and storage of the toxic metal is already known from hyperaccumulator plants. It usually results in metal storage in organs and tissues where it interferes least with the sensitive metabolic processes like photosynthesis and respiration, i.e. the vein and the epidermis.
metal-chelating ligands phytochelatins (PCs) were detected in extracts from the plants. The induction of different PCs was not proportional to the applied Cd concentration, but occurred in a switch-like manner and specifically for each PC species. The most noticeable increase was PC3 at the threshold concentration of 20 nM Cd.
A combination of different heavy metals and other factors caused the nearly complete lack of macrophytes in an oligotrophic hard water lake (lake Ammelshain). Within the lake, elevated concentrations of Cd (3 nM), Nickel (300 nM) and reduced Phosphate (75 nM) seemed to be responsible for the lack of aquatic plants and were tested for their inhibitory capability in hard and soft water. While the single treatments with non-toxic concentrations of Cd or slightly toxic concentrations of Ni caused no or only minimal toxicity symptoms in C. demersum, they became highly toxic when applied in combination. This negative effect was even more severe under phosphate-limitation. High concentrations of Calcium and Magnesium in the lake water reduced metal toxicity, indicating additional reasons for the absence of macrophytes in the lake. But regarding other freshwater habitats, these measurements revealed that synergistic metal toxicity may be an important influencing factor for the colonisation of soft waters by water plants.
Altogether, the results from this thesis indicate the onset of Cd toxicity and detoxification in a model water plant at a significantly lower level than shown in previous studies.
Zusammenfassung
Diese Arbeit befasst sich mit den Effekten, die das hochtoxische Schwermetall Cadmium auf die wurzellose, aquatische Modellpflanze Ceratophyllum demersum ausübt, welche sowohl biochemisch als auch biophysikalisch ermittelt wurden. Dabei wurden die Stressexperimente unter umweltrelevanten Bedingungen durchgeführt. Dies bedeutet, dass Licht und Temperatur über die Zeit einer sinusförmigen Kurve folgten, dass das Verhältnis von Biomasse zu Wasservolumen in den Aquarien so niedrig war, wie es in einem nährstoffarmen See der Fall ist, und dass frisches Nährmedium ständig zugeführt wurde, damit die Metallaufnahme in die Pflanze nur durch die Aufnahmekapazität der Pflanze, nicht aber durch die Verfügbarkeit limitiert war. Außerdem wurden Cadmiumkonzentrationen im nanomolaren Bereich verwendet und die Behandlungsdauer war lang genug gewählt, um auch chronische Toxizität beobachten zu können.
Durch die Toxizitätsstudie zeigte sich, dass der Photosyntheseapparat der erste Ort der Inhibierung ist. Die maximale Quantenausbeute von Photosystem II (PS II) im dunkeladaptierten Zustand und der Elektronenfluss durch das PS II im aktinischen Licht wurden stark reduziert. Das Auftreten reaktiver Sauerstoffspezies (ROS) wurde erst nach längerer Behandlungsdauer beobachtet, was darauf hinweist, dass ihre Bildung nicht der Grund, sondern die Folge der nicht-funktionalen Photosynthese ist. Für die meisten der gemessenen Parameter wurde eine Grenzwertkonzentration von 20 nM Cd bestimmt. Ab dieser Konzentration wurde das durch Cadmium ausgelöste verstärkte bzw. verringerte Auftreten eines Effekts beobachtet. Dies ist eine viel geringere Konzentration als in vielen vorherigen Studien eingesetzt wurde. Alle beobachteten Effekte waren stärker ausgeprägt in Pflanzen, die bei hoher (verglichen mit geringer) Lichtintensität unter Cd-Stress standen.
Dieser Umstand deutet auf eine schützende Rolle des in Schwachlicht viel stärker ausgeprägten Antennensystems der Pflanzen hin.
Weiterhin führte die Behandlung mit Cadmium zu einer Umverteilung von anderen Metallen im Gewebe der Versuchspflanze, vor allem von Zink, wie mit Hilfe der nichtinvasiven Röntgenfluoreszenz (µ-XRF) an schockgefrorenen Blättern festgestellt werden konnte. Während bei geringer und mittlerer Cadmiumkonzentration Zink hauptsächlich in der Epidermis und im Mesophyll der Blätter gefunden wurde, fand bei hochtoxischer Cadmiumkonzentration eine Umverteilung im Gewebe statt. Zink wurde nun vor allem im Leitbündel gefunden, vermutlich weil die Zinktransporter durch Cadmium blockiert waren.
Eine Umverteilung von Cadmium bei steigender Konzentration wurde als Teil des Detoxifizierungsprozesses beobachtet. Während kein Cadmium in den Kontrollpflanzen (keine Cadmiumbehandlung) detektiert wurde, konnte bei geringer Cadmiumkonzentration eine homogene Verteilung über den gesamten Blattquerschnitt beobachtet werden. Bei moderat toxischen, sowie hochtoxischen Konzentrationen fand eine Einlagerung von Cadmium in diejenigen Gewebe statt, in denen Cadmium am wenigsten in sensitive, metabolische Prozesse wie Photosynthese und Atmung eingreifen kann. Diese Gewebe sind vor allem das Leitbündel und die Epidermis, wie bereits von Hyperakkumulatorpflanzen bekannt ist.
Als weiterer Entgiftungsmechanismus trat die Induktion der metallbindenden Liganden Phytochelatine (PCs) auf. Die Induzierung der einzelnen PC-Arten war dabei jedoch nicht proportional zur verwendeten Cadmiumkonzentration. Vielmehr gab es für jede PC-Art eine spezifische Grenzwertkonzentration, bei der eine erhöhte Akkumulation
gemessen werden konnte. Der auffälligste Anstieg fand für PC3 bei einer Konzentration von 20 nM Cadmium statt.
Eine Kombination verschiedener Schwermetalle sowie anderer Faktoren waren für das beinahe vollständige Fehlen der Makrophytenflora in einem nährstoffarmen See (Ammelshainer See) mit hoher Wasserhärte verantwortlich. In besagtem See waren vermutlich erhöhte Konzentrationen von Cadmium (3 nM) und Nickel (300 nM) bei niedrigem Phosphatgehalt (75 nM) der Grund für das Fehlen der Makrophyten. Daher wurden die Metalle einzeln und in Kombination in Wasser mit geringer und hoher Härte auf ihre inhibitorische Wirkung hin getestet. Einzeln eingesetzt zeigten sich keine (Cadmium) oder nur geringe (Nickel) Stresssymptome. In Verbindung, und vor allem mit zusätzlicher Phosphatlimitation waren die Effekte ungleich schwerwiegender. Im See sorgten allerdings die hohe Wasserhärte, also die erhöhten Konzentrationen von Magnesium und Calcium, für geringere Metalltoxizität, so dass noch weitere Parameter für das Fehlen der Makrophytenflora verantwortlich sein müssen. Allerdings zeigen die Ergebnisse dieser Feldstudie, dass synergistische Hemmung durch die Toxizität mehrerer Metalle eine entscheidende Rolle für die Besiedlung von Gewässern durch Wasserpflanzen haben kann.
Die Ergebnisse dieser Arbeit zeigen, dass sowohl Toxizität als auch Detoxifizierung von Cd in der aquatischen Modellpflanze bei weit geringeren Konzentrationen auftreten, als bisher angenommen.
Abbreviations
AAS – Atomic Absorption Spectroscopy APX – Ascorbate peroxidase
CA – Carbonic anhydrase CAT – Catalase
CEBAS – closed equilibrated biological aquatic system Chl – Chlorophyll
CRL – compound refractive lenses DESY – Deutsches Elektron Synchrotron DNA – Deoxyribonucleic acid
dw – dry weight
EDDHA - Ethylenediamine-N,N’-bis (2-hydroxyphenylacetic acid) EDTA - Ethylenediaminetetraacetic acid
ESI – Electrospray ionisation
FKM – Fluorescence Kinetic Microscope fw – fresh weight
GSH – Glutathione
Hms – Heavy metal substituted LHC – Light harvesting complex ICP – Inductively coupled plasma IRT – Iron regulated transporter MN – Micronucleus
MS - Mass spectrometry MT - Metallothionein
NADPH - Nicotinamide adenine dinucleotide phosphate
OMEGAHAB – Oreochromis-mossambicus-Euglena-gracilis-aquatic-habitat PC – Phytochelatin
PETRA - Positron-Elektron-Tandem-Ring-Anlage PF2 - Peroxyfluor-2
ppm – parts per million, mg kg-1 or µg g-1 PS - Photosystem
ROS – Reactive oxygen species
RubisCO – Ribulose-1,5-bisphosphate carboxylase/oxygenase SMNS – submerged macrophyte nutrient solution
SOD – Superoxide dismutase UV/Vis – ultra violet / visible light ZIP – Zinc – IRT-like protein ZnT – Zinc transporter
1. Introduction
1.1. Heavy metals in plants
Heavy metals play an ambiguous role in plant nutrition. While some metals like copper (Cu) or zinc (Zn) are essential for normal growth and development of plants, others are only toxic to them (e.g. mercury, Hg). However, even essential elements can become toxic when the concentrations exceed optimal levels. The heavy metal cadmium (Cd) is very toxic for all organisms with only very specific exceptions. In this thesis, the effects of nanomolar Cd concentrations on water plants were investigated and the mechanism of its toxicity determined.
In the following, an introduction to Cd itself and its toxicity in plants will be given.
Afterwards, the system in which the toxicity mechanisms were determined in this thesis will be presented.
1.2. Cadmium
Cadmium (Cd) is a chemical element with the atomic number 48, in group 12 of the periodic system, also referred to the zinc group. The zinc group also contains the essential element zinc (Zn) and the toxic element mercury (Hg), with both of them Cd shares certain chemical properties. Cadmium is mostly associated with Zn, and indeed in 1817 Cd was discovered as an impurity of zinc carbonate (ICdA, 2013). Cadmium ranges in the middle of the heavy metals regarding its density (8.46 g cm-3 at room temperature, RT) and atomic mass (112.4 u or g mol-1). Cadmium is a rather rare element in the Earth’s crust with mean values between 0.1 and 0.5 parts per million (ppm) and holds the 64th place of naturally occurring elements regarding the average abundance (ICdA, 2013).
Natural occurrences of Cd are linked to volcanic eruptions, burning of vegetation, and chemical weathering of rock material (Cullen and Maldonado, 2013). A much higher contribution of Cd release into the environment results from anthropogenic actions. As Cd is very often associated with Zn, it is a common by-product from Zn ores (Chizhikov, 1966;
Nriagu, 1980). With the industrial revolution, the use of Cd increased steadily (recently reviewed by Cullen and Maldonado, 2013). Cadmium is found, e.g., in plastic stabilizers used
in car tires, as a yellow pigment in paint, inks, rubbers, as a component in rechargeable Nickel-Cadmium batteries, and in tobacco smoke.
Cd is very mobile in soils and particles dissolve easily in water (unlike e.g. Fe, which is insoluble and biologically unavailable in the form of Fe3+). Concentrations vary from pM (10-12 mol l-1) in the oceans (Chung and Pai, 1996) via nM (10-9 mol l-1) in uncontaminated waters like Lake Constance (Petri, 2006) to µM (10-6 mol l-1) in stream Naraguta in Nigeria, which is heavily contaminated by tanning activities (Ahmed, 2011). The high release into the environment is a great threat because Cd is highly toxic to all organisms. Some important pollution scenarios were linked to Cd poisoning, like the “itai-itai-disease” in Japan in the early 20th century (Kaji, 2012). The effects of Cd on humans are very well studied nowadays.
Acute or chronic toxicity can lead to multiple effects ranging from salivation, nausea, to organ (especially kidney) failure, nephrotoxicity and cancer (recently reviewed by Hartwig, 2013;
Thévenod and Lee, 2013). Besides inhalation and smoking, the main entry route of Cd into the human body is food. Plants take up Cd easily, and plant food averagely contains higher Cd concentrations than meat, eggs, or milk products (Thévenod and Lee, 2013) and allows Cd to enter the human food chain (McLaughlin et al., 1999). Therefore, the irrigation of crop plants used either directly for vegetable production, or for the production of animal feed with municipal waste water or the fertilization with sewage sludge or igneous phosphate fertilizers (Gimeno-Garcia et al., 1996; Thévenod and Lee, 2013) pose a great risk for human health (Kalavrouziotis et al., 2009).
1.2.1. Cadmium in plants
1.2.1.1. Cadmium in phytoplankton
Cd is also highly toxic for photosynthetic organisms. One exception has been found in which Cd has a metabolic function under natural conditions. The first observation was that the vertical distribution profile of Cd in the oceans resembles the one of the phytoplankton nutrient phosphate: While the concentrations are depleted at the surface water due to uptake by phytoplankton, they increase in deep water due to subsequent remineralization of sinking organic matter with a maximum in the main thermocline (Boyle et al., 1976; Bruland et al., 1978).
Indeed, a metabolic function was found in Zinc-depleted cultures of the marine diatom
low concentrations of Cd (Lane and Morel, 2000). The growth enhancement was also observed for other species of marine phytoplankton (summarized in Xu and Morel, 2013), but only when Zn was reduced and not completely removed (Xu et al., 2007). Furthermore, the growth enhancement by Cd under Zn-limited conditions was always lower than without Cd under Zn-sufficient conditions. The reason for the positive effect of Cd was identified in the form of the enzyme carbonic anhydrase (CA). The CA belongs to a large protein family and catalyses the conversion of hydrogen bicarbonate (HCO3-
) into carbon dioxide (CO2) and vice versa. In plants and algae, CO2 is fixed by the enzyme ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO). The conversion of HCO3-
to CO2 enriches the concentration of CO2 in close proximity of RubisCO and thereby suppresses the oxygenase activity of the enzyme, and as a consequence, photorespiration. The activity of the CA is therefore important for carbon fixation.
Diatoms express two different kinds of CAs, depending on the metal availability in the oceans: Under conditions where Zn is not limiting, diatoms possess the δ-form of the CA (Lane and Morel, 2000). This form contains either a Zn or a Cobalt (Co) ion in its active center. Under Zn limitation, however, a protein is expressed that contains either Cd or Zn (ζ-CA). Although the catalytic activity of the ζ-CA is higher when Zn is bound instead of Cd, the substitution allows the diatoms to cope with Zn limitation, yield fast growth rates and therefore are able to compete with other algae in the low metal environment of the oceans (Lane and Morel 2000; Xu et al. 2008). However, the cdca gene coding for the ζ-CA was found in some (like Thalassiosira pseudonana), but not all other species (like Emiliania huxleyi or Tetraselmis maculata, Lee and Morel, 1995; Xu et al., 2007) that showed a growth enhancement when supplemented with low Cd concentrations. In those species, Cd must have another positive role for other, yet unknown reasons.
Whether Cd has a positive effect in land plants is still to be determined. However, there is a strong hint that also in the hyperaccumulator Noccaea (formerly Thlaspi) caerulescens Cd is used in the active center of the CA. When Cd was added to the medium, growth was enhanced and the enzyme was shown to be more active (Liu et al., 2008) compared to non- supplemented media.
1.2.1.2. Hyperaccumulators
As plants are immobile and cannot avoid unfavorable conditions, like high concentrations of heavy metals, they have to adjust and adapt to them. Some plants actively accumulate large amounts of heavy metals and can therefore grow on polluted soils. Those
“hyperaccumulators” do not take the metals up for metabolic reasons, but store them in tissues and organs (mostly cell wall and vacuole) where they interfere least with the metabolic processes (Küpper and Leitenmaier, 2013). Besides occupying an ecological niche, the hyperaccumulators also gain advantage over related non-accumulators as they are protected from herbivores or pathogens. This “defense hypothesis” was verified for many plants, metals and pathogens. A summary of the available data including experimental approaches and other challenges when studying the elemental defense hypothesis was released by Boyd (2007).
1.2.2. Cd toxicity in plants
Beside the two mentioned scenarios of positive effects, there is a vast number of studies emphasizing the toxicity of Cd (recently reviewed by Andresen and Küpper, 2013). Some of the main threats that also concern the following specialized chapters will be summarized below.
1.2.2.1. Damage to the photosynthetic apparatus
For phototrophic organisms the inhibition of photosynthesis is certainly a main threat. Cd can inhibit photosynthesis by direct interaction with the photosynthesis-related molecules. The substitution of Mg2+ in chlorophyll (Chl) by different divalent metal ions was shown for different plant and algae species (Kowalewska et al., 1992, 1987; Kowalewska and Hoffmann, 1989; Küpper et al., 1996, 1998, 2002, 2006). Those heavy-metal substituted Chls ([hms]-Chls) are unsuitable for photosynthesis for many reasons. The singlet excited state of [Cd]-Chl is very unstable and the return to the ground state is achieved by thermal relaxation rather than fluorescence or electron transfer (= photosynthesis) (Watanabe and Kobayashi, 1988; Watanabe et al., 1985). The capacity of electron release from the excited state is highest for [Mg]-Chl compared to any other [hms]-Chl (Watanabe et al., 1985). And many [hms]-Chls cannot bind axial ligands (Boucher and Katz, 1967), making the structure of the
Depending on the light conditions, the substitution of Mg2+ by other heavy metals takes place in different molecules: Under low light conditions, it will be in the LHC molecules forming the antenna system (shade reaction), while under high light conditions the replacement will be directly in the reaction centers (RC; sun reaction) damaging the PS II core (Küpper et al., 2002). Although these reactions are well studied for e.g. Cu, in vivo proof is lacking for [Cd]- Chl, which is hard to detect, as its absorption spectrum resembles that of [Mg]-Chl and cannot be distinguished by UV-Vis spectroscopy. However, [Cd]-Chl is rather unstable (unlike [Cu]-Chl) and easily degraded (Küpper et al., 1996). Only a small fraction (3-10%) of [Mg]-Chl will be substituted by Cd under environmentally relevant conditions (see below).
And although only these few Chl substitutions can be tremendous for the plant, especially under high light conditions, their detection becomes a challenge. However, inhibition of photosynthesis due to Cd treatment was clearly observed as reduced electron flow through PS II, reduced variable fluorescence, or enhanced non-photochemical fluorescence quenching (Küpper et al., 2007a; Pietrini et al., 2010) suggesting the occurrence of the substitution.
1.2.2.2. Induction of reactive oxygen species
Oxidative stress due to heavy metal treatment is a well studied phenomenon in plants and algae (reviewed by Pinto et al., 2003; Shaw et al., 2004). Reactive oxygen species (ROS) are a normal by-product of metabolic processes (respiration and photosynthesis; Asada 2006;
Asada and Takahashi, 1987) and also occur in specialized organelles like peroxisomes (Corpas et al., 2001; del Río et al., 2002). Low concentrations of e.g. nitric oxide (NO) and hydrogen peroxide (H2O2) serve as signaling molecules in the cell (reviewed by Mittler et al., 2011; Van Breusegem et al., 2008). Plants and other phototrophic organisms possess several antioxidant enzymes and scavenging molecules to dispose off the harmful ROS. While under unstressed conditions, the formation and neutralization of ROS is balanced, under stress conditions, the imbalance is in favor of the accumulation of ROS, leading to the oxidative stress. The two most reactive species are the hydroxyl radical (HO•) and singlet oxygen (1O2).
They are considered to be specifically dangerous to all living cells as they react fast and non- specific with all kinds of biomolecules, including DNA, proteins, enzymes, and lipids (Shaw et al., 2004). The peroxidation of unsaturated fatty acids of plasma- and organelle membranes can lead to their disruption and thereby leakage of the organelle and cell contents. One further danger of lipid peroxidation is the production of mutagenic aldehydes (Halliwell and Gutteridge, 2007).
As Cd is redox-inert and does not react directly with molecular oxygen, the most likely origin of ROS under Cd stress is the dysfunction of photosynthesis and respiration. Increased concentrations of superoxide radicals (O2•-
) and hydrogen peroxide (H2O2) were observed in Cd-treated pea plants by specific dyes (Romero-Puertas et al., 2004; chapter 2.1. of this thesis). Both, the enhancement of ROS production and the reduction of enzymatic activity of the antioxidant system were connected to Cd toxicity. An upregulation of gene expression of those enzymes is clearly a response to enhanced oxidative stress. The reduction in activity, e.g. of the superoxide dismutase (SOD) may be the result of the substitution of the Zn ion in its active center by Cd (see below), thereby inactivating the enzyme (Sandalio et al., 2001) which will lead to higher amounts of ROS.
1.2.2.3. Cadmium-induced genotoxicity
Cadmium can induce genotoxicity by directly interacting with the nucleotides (Müller, 2010;
Sigel et al., 2013), by inhibiting DNA-repairing enzymes, and via the induction of ROS leading to lipid peroxidation, membrane leakage, and the production of mutagenic aldehydes (Lin et al., 2007). In Vicia species, the root tip micronucleus assay is well established (Ma et al., 1995). Micronuclei (MN) originate during cell division by the exclusion of whole chromosomes or by chromatin fragmentation (Savage, 2000). In studies with Allium cepa, Allium sativum, and Vicia faba, treated with different Cd concentrations from 75 nM up to 200 µM, the MN formation was significantly higher in the roots treated with Cd compared to the control (Manier et al., 2009; Seth et al., 2008; Ünyayar et al., 2006). Cd-induced DNA damage was also shown as increasing length of tail DNA in the so called Comet Assay. In this assay, nuclei of Cd-treated cells were isolated, lysed and damaged DNA was determined by the amount that left the nucleus when gel electrophoresis was applied. The higher the damage, the longer will be the tail of the comet (Cvjetko et al., 2010; Koppen and Verschaeve, 1996).
Cadmium-induced damage to plants can have several reasons and can reveal itself in many ways. Most of the proposed mechanisms are summarized in the following graph.
Figure 1: Scheme of damage pathways and interactions of Cd toxicity which were proposed in the literature. (Picture taken from Andresen and Küpper, 2013).
1.3. Environmentally relevant experiments
One major disadvantage of many previous studies, which also led to the above mentioned conclusions about Cd toxicity, is the high concentrations which were applied. No higher concentrations than a few µM were ever found in natural habitats, and these were often the result of contamination nearby (Ahmed et al., 2011; ICdA, 2013). It is questionable whether the observed effects of experiments using many µM to mM concentrations of Cd (Garnier et al., 2006; Iannone et al., 2010; Sigfridsson et al., 2004) would occur also in plants growing in Cd-polluted areas. Though these studies underline the toxicity of Cd, they do not allow insight into different toxicity mechanisms, including determination of threshold concentrations for measured effects on photosynthesis, respiration, or ROS production. Furthermore, many nutrient solutions in previous studies contained EDTA (ethylenediaminetetraacetic acid), a metal-chelating agent. It is used to prevent iron-phosphate precipitation and thereby ensuring
iron bioavailability (Dalton et al., 1983) as iron is redox active and in form of Fe3+ insoluble and biologically unusable. But on the other hand, EDTA binds other, mostly divalent metal ions, thereby reducing their availability and therefore toxicity of e.g. Cd. This means that even if high concentrations of Cd were used, by having EDTA in the nutrient solution the available amount of Cd for the plant was not known.
1.4. Setting of the experiments in this thesis
The experiments undertaken in this thesis investigated the mechanism of Cd toxicity under controlled, environmentally relevant conditions. This means in particular that concentrations in the nM range were applied and the nutrient solution contained a specific iron chelating agent (ethylenediamine-N,N’-bis (2-hydroxyphenylacetic acid), Fe-Eddha; Brown et al., 1960). As the aquatic plant Ceratophyllum demersum L. was used, low biomasses to water ratios as occurring in oligotrophic lakes were ensured. Furthermore, fresh nutrient solution was continuously added to the aquaria with an overflow limiting the volume to 2 L (chemostatic culture conditions) and light and temperature followed a sinusoidal cycle to simulate night, dawn, midday and dusk. Each experiment lasted at least 2 weeks to simulate chronic toxicity. Measurements of e.g. physiological parameters were carried out continuously during the treatment to follow the effects of Cd and to determine not only concentration-, but also time-dependent thresholds of Cd toxicity.
1.4.1. Ceratophyllum demersum – a free floating, rootless, submerged macrophyte
For all experiments of this study, the submerged aquatic macrophyte Ceratophyllum demersum L. was used. Aquatic plants are useful for studying effects of metal toxicity. A hydroponic solution allows the application of a defined nutrient composition with knowledge of all components, which is not the case when soil substrate is involved. Also, the application of specific heavy metals in form of salts is very easy, as the salts dissolve easily in water.
The use of a rootless, submerged plant is especially insightful to study toxicity mechanisms occurring specifically in the photosynthetically active shoot. In terrestrial plants,
roots are by far the most important organs for the uptake of metals and non-metal nutrients.
Therefore, roots are the first organs to be affected by Cd toxicity. Effects of Cd on the shoots of terrestrial plants could also be the consequence of Cd-induced damage to the roots. On the other hand, roots are also the first barrier against Cd. Defense mechanisms like enhanced lignification (Ederli et al., 2004) or impregnation of the cell walls with suberin (Lux et al., 2011) have been observed. These root-specific mechanisms which in principle can also affect reactions in the shoot are not possible when using a root-less plant. C. demersum takes up all nutrients directly via the leaves and the stem; it is free floating and does not need any solid substrate and has therefore already been used in tests of biological life support systems on space flights (Blüm et al., 1994; the CEBAS and OMEGAHAB systems). Ceratophyllum is a cosmopolitan genus and can be found in lakes or slow-moving rivers with sufficient nutritious levels on all continents except Antarctica (GISD, 2006).
1.5. Fluorescence kinetic measurements
When photosynthetic organisms absorb light with their antenna molecules, there are three major ways of Chl de-excitation: photochemistry, heat dissipation and fluorescence. As all three possibilities compete with each other, the measurement of one can give information about the others. Accordingly, Chl fluorescence has been used to investigate the photosynthetic activity of plants for many decades (Baker, 2008; DeEll and Toivonen, 2003;
Kautsky and Hirsch, 1931; Maxwell and Johnson, 2000). Most measuring protocols follow this established sequence:
First, the leaf is allowed to adapt to darkness to convert the plastoquinone pool into oxidized state. That means, any photon absorbed by the antenna will drive photosynthesis (the PS II RC is “open”) and therefore the fluorescence will be minimal (F0). When a saturating light flash is applied, the plastoquinone pool gets reduced within ms, there is no possibility to accept another electron from PS II (the PS II RC is “closed”) and the energy will be released by fluorescence and will not be used to drive photosynthesis, i.e. fluorescence is maximal (Fm). The difference between Fm and F0 is Fv, the variable fluorescence; standardized to Fm, the values of Fv/Fm display the maximal photochemical quantum yield of the PS II RC in dark adapted state. Although changes in Fv/Fm can be the result of many factors, it is widely used as a general stress parameter (Baker, 2008).
In a second phase, the leaf gets illuminated with actinic light, which promotes photosynthesis and thereby reduces fluorescence. Saturating light flashes allow the determination of Fm’ (maximal fluorescence in actinic light). The electron flow through PS II at any given time point t can be determined as the difference between Fm’ and the fluorescence at this time point divided by the maximal fluorescence in actinic light: ΦPSII or qP = (Fm’-Ft)/Fm’ (Genty et al., 1989). As fluorescence gets reduced due to performance of photochemistry, ΦPSII is termed
“photochemical (fluorescence) quenching”.
The respective contrast of ΦPSII is the reduction of fluorescence due to upregulation of heat dissipation and termed “non-photochemical (fluorescence) quenching” (qNP; NPQ). It is calculated from (F m/Fm’)-1.
In the third phase, the actinic light is switched off again and the ability of the leaf to return to the state of dark adaptation can be observed.
Three kinds of lights are used in a fluorescence measurement. Pulse measuring light consits of ultrashort pulses (10-250 µs), which integrated over time are of too low intensity (< 2.5 µmol m-2 s-1) to reduce plastoquinone and drive photosynthesis; it, is used to measure F0. Saturating light pulses are several hundred ms (here: 600 ms) and of high intensity (3200 µmol m-2 s-1); they immediately reduce the plastoquinone pool and close the PS II RC.
Actinic light (350 µmol m-2 s-1) is used to drive photosynthesis.
In this thesis, Chl fluorescence was determined using the Fluorescence Kinetic Microscope (FKM, Küpper et al., 2000a, 2007a). The setup does not only allow measuring the above described parameters, the addition of a camera synchronized high-sensitive fibreoptic spectrometer enables us to record absorption and fluorescence spectra of the measured leaf and observe fluorescence changes at specific wavelengths. For instance, state transition, the movement of LHC molecules from PS II to PS I or vice versa could be observed, as was done for the phycobiliproteids in the cyanobacterium Trichodesmium erythraeum ISM101 (Küpper et al., 2009).
Figure 2: Example of the measuring protocol for Chl a fluorescence at the FKM. A dark adapted leaf is subjected to light of different qualities and the resulting fluorescence is recorded. Black arrows represent saturating light flashes. There are six flashes during the actinic light phase (i1 to i6) and five in relaxation phase (r1 to r5). See text for explanations.
1.6. Aims of this study
The effect of Cd on plants is manifold and strongly dependent on Cd concentration and the exposure duration. Many proposed mechanisms were obtained using rather unnatural conditions with respect to experimental setup, applied concentrations and period of observation. It is important to investigate the mechanisms of Cd-induced damage under environmentally relevant conditions to judge the endangerment of Cd toxicity in nature.
Therefore, the aims of this thesis were:
The determination of threshold concentrations and time points of the manifestation of Cd toxicity in water plants and the mechanisms behind it (chapter 2.1.). This study focused on the physiological response and the effect on the overall photosynthetic performance of C. demersum.
The investigation of detoxification mechanisms and their concentration-dependent change in Cd-treated plants (chapter 2.2.).
The correlation of the laboratory-obtained results with an actual environmental problem (chapter 2.3).
2. Publications in peer-reviewed journals and manuscripts
2.1. Cadmium toxicity investigated on physiological and biophysical level under environmentally relevant conditions using the aquatic model plant Ceratophyllum demersum L.
Elisa Andresen1, Sophie Kroenlein1, Hans-Joachim Stärk2 and Hendrik Küpper1,3
1) University of Konstanz, Department of Biology, D-78457 Konstanz, Germany
2) UFZ – Helmholtz Centre for Environmental Research, Department of Analytical Chemistry, Permoserstr. 15, D-04318 Leipzig, Germany
3) University of South Bohemia, Faculty of Biological Sciences and Institute of Physical Biology, Branišovská 31, CZ-370 05 České Budejovice, Czech Republic
Unpublished manuscript
Abstract
Cadmium is an important environmental pollutant and poisonous to most organisms. We investigated the mechanisms of Cd toxicity in the aquatic, rootless model plant Ceratophyllum demersum using environmentally relevant conditions including nature-like light and temperature cycles and Cd concentrations in the nM range. As expected, the level of inhibition increased with increasing Cd concentrations. The threshold concentration for most parameters was 20 nM, below, hardly any stress symptoms were observed. The first site of inhibition was photosynthesis (measured as Fv/Fm, ΦPS II), followed by increased production of reactive oxygen species, most likely a follow reaction of dysfunctional photosynthesis and energy dissipation. Cd treatment induced changes in pigment contents, reducing Chl and increasing quenching pigments. All observed effects were more pronounced in plants cultivated under high light conditions compared to low light conditions.
Introduction
The heavy metal cadmium is an important environmental pollutant and toxic to most organisms. Concentrations in the Earth’s crust are rather low (0.1-0.5 ppm, Maret and Moulis, 2013), but mining, smelting and industrial use (e.g. as plastic stabilzer, in pigments and in NiCd batteries) still increase the amount of Cd in the environment. Cd is highly water soluble and can be easily taken up by plants, thus entering the human food chain (McLaughlin et al., 1999). Due to chemical similarity with Zn, Cd is mainly taken up via transporters with similar affinity for Cd and Zn, and many toxic effects are related to Zn replacement or limitation (Clemens, 2006). Sewage sludge and organic fertilizers contaminated with Cd bring it directly into contact with crop plants and should be avoided for this reason (He et al., 2005; McBride et al., 1997).
Under specific circumstances, Cd was found to have a metabolic function. The first observation was that diatoms under severe Zn-limitation show enhanced growth, when Cd was added (Price and Morel, 1990). Later experiments revealed that those diatoms express a carbonic anhydrase (CA), which can have Cd in its active center, when Zn is not available (CDCA, ζ-CA). Growth enhancement under Zn-limitation was also observed in other species of marine phytoplankton, which do not possess the cdca gene (recently reviewed by Xu and Morel, 2013), suggesting another biochemical role of Cd, which could explain the micronutrient-like distribution of Cd in the oceans (Pai and Chen, 1994).
Besides the above mentioned positive function, there is overwhelming evidence for cadmium toxicity in plants (Andresen and Küpper, 2013). Cadmium can interfere with all parts of a plants’ metabolism: The uptake or translocation of other, essential nutrients can be inhibited, because Cd competes with other ions for the binding sites (Dong et al., 2006), making deficiency or imbalance of essential nutrients a part of Cd toxicity. Not only the binding sites of transporters can be filled by Cd, other ions in enzymes can be replaced by Cd as well. The substitution in a structural site can, but also may not result in great changes in the protein structure. But if the substitution occurs in the active center, it will make the enzyme inoperable, as redoxreactions are not possible with redox-inert Cd (Maret and Moulis, 2013).
A great threat for all photosynthetic organisms is the inhibition of photosynthesis. It was shown that Cd treatment led to decreased photosynthetic activity in various plant species (see reviews by Andresen and Küpper, 2013; Küpper and Kroneck, 2005) with a much stronger inhibition of PS II compared to PS I (Atal et al., 1991; Clijster and Van Assche, 1985). The replacement of the Mg2+-ion in the chlorophylls of the reaction centers and in the light harvesting complexes (LHCs) is one result of heavy-metal treatment in photosynthetic
organisms. The heavy-metal substituted chlorophylls ([hms]-Chls) are unsuitable for photosynthesis (Kowalewska et al., 1992, 1987, Kowalewska and Hoffmann, 1989). The mode of substitution is strongly dependent on the irradiance as shown by Küpper et al. (1998, 2002), leading to stronger inhibition of photosynthesis by heavy metals in high light (Cedeno- Maldonado and Swader, 1972). Cadmium can enhance the presence of reactive oxygen species (ROS) in plants and algae (reviewed by Andresen and Küpper, 2013; Sandalio et al., 2009; Pinto, 2003). As Cd is a redox-inert metal, ROS cannot be produced directly via Fenton or Haber Weiss reaction (Wardman and Candeias, 1996), but indirectly by impairment of respiration and photosynthesis leading to the mis-transfer of electrons to oxygen instead of their target molecule. Alternatively, the antioxidative system can be inhibited by Cd, resulting in a reduced removal of existing ROS (Sandalio et al., 2001).
All of these effects have been observed in plants, but in most studies, very high, environmentally unrealistic Cd concentrations (upto mM, e.g. Sigfridsson et al., 2004) were used. Cd concentrations in unpolluted environments range from 0.2-0.4 nM (e.g. Lake Constance: Zweckverband Bodensee-Wasserversorgung, 2011; Petri, 2006) to 5 nM in slightly polluted rivers in Germany (Bachor et al., 2012). Even in a heavily polluted stream in Nigeria, the highest concentration measured was 1.3 µM (Ahmed et al., 2011). Effects observed under very high Cd concentrations are therefore questionable as plants would never be exposed to them in nature. The effects may be very different under acute (hours to days) vs. chronic (days to months) toxicity. Additionally, laboratory conditions simulating natural conditions (e.g. light and temperature cycles apart from switch on / switch off) may also lead to new insights into toxicity mechanisms. Many studies show that Cd is toxic to plants, that photosynthesis is inhibited, or that Cd treatment induces reactive oxygen species, but the causal relationship often remained unclear.
This study aimed to fill this gap and investigate the biochemical and biophysical mechanisms of Cd toxicity under environmentally relevant conditions. We used nature-like light and temperature conditions and applied Cd concentrations from less than 1 nM to 200 nM.
Material and methods
Plant material and culture conditions
The rootless, submerged and free-floating macrophyte Ceratophyllum demersum L. was used for the stress experiments. The strain was obtained from an aquaria shop and continuously cultivated since 2005 in hydroponic solution with 12 h day/12 h night light conditions provided by two Osram FLUORA® fluorescent and two warm white fluorescent tubes (Osram, München, Germany) and a temperature cycle from 18°C at 6 a.m., over 20°C at 9 a.m., to a maximum of 22°C at 3 p.m., back over 20°C at 9 p.m. to 18°C again at 6 a.m. The nutrient solution (SMNS, submerged macrophyte nutrient solution) was optimized for growth of submerged macrophytes and resembled the situation of typical oligotrophic waters, in particular soft waters (Andresen et al., 2013b). The pH was adjusted to 7.8 with KOH. All experiments were carried out under simulations of natural light and temperature conditions:
12 h night and 12 h sinusoidal light cycle with maximal irradiances at 500 µmol photons m-2 s-1 (supplied by Dulux L 55 W / 12_954, OSRAM München, Germany) for high light experiments and 60 µmol photons m-2 s-1 for low light experiments. The temperature cycle was 19°C at 6 a.m., 21.5°C at 9 a.m., 24°C at 3 p.m., 23 C at 9 p.m., 19 C at 6 a.m. For each treatment around 1.5 g (fresh weight) of plants were used. The number of individual plants was consistent for each concentration within the same experiment. Differences occurred due to weight and size of the plants at treatment start. Each aquarium contained 2 L of medium to secure a low biomass to water volume ratio, which was also constantly aerated by room air. The nutrient solution was continuously exchanged (flow rate 0.5 L day-1) to ensure that the metal uptake into the plants was limited only by the concentration, but not by the amount of nutrient solution available. After at least one week of acclimation to the experimental light conditions, cadmium was applied as CdSO4 to the medium. The concentrations were 0.2, (background, no Cd added), 0.5 (LL only), 1, 2, 5, 10, 20, 50, 100, and 200 nM. Epiphytic algae and cyanobacteria were weekly removed from the macrophytes by gentle brushing in SMNS (without micronutrients) and the aquaria cleaned with ddH2O.
Three experiments were carried out under low light (LL) conditions, two under high light (HL) conditions. A third HL experiment was performed, in which the plants showed the same trends of inhibitions, until they died after 5 weeks for unknown reasons.
All of the described measurements of physiological parameters were done using the nutrient solution SMNS without micronutrients (Rochetta and Küpper, 2009).
Fluorescence kinetic microscopy
Physiological changes in the plants induced by cadmium treatment were determined by two- dimensional (imaging) microscopic measurements using the Chl fluorescence kinetic microscope (Küpper et al., 2000a, 2007a). One leaf from the 5th nodium counted from the apex of the plant was fixed in the measuring chamber by gas-permeable cellophane and the area just before the branching (approximate size of 1.1 x 1.1 mm) used for the spatially and spectrally resolved kinetic measurements. A continuous flow of tempered (25°C) culture medium kept the sample under physiological conditions. A detailed description of the microscope and the used protocols of the Kautsky induction can be found in Küpper et al.
(2007a) and Andresen et al. (2010). These measurements were done weekly, one day after the cleaning. The leaves from the FKM measurement were frozen in liquid nitrogen and stored at -80°C after use for pigment analyses.
Oxygen exchange
After 6 weeks of treatment net photosynthetic oxygen release and respiratory oxygen uptake were measured with a Clark-type electrode (CellOx® 325, OxiCal-SL, WTW, Weilheim, Germany) in a custom-made 200 ml measuring chamber at 25°C. Oxygen uptake was measured in the dark, and oxygen release under increasing light intensities (from darkness to 466 µmol photons m-2 s-1). Data were recorded and analyzed using the OxyCorder measuring device with the software Oxywin 3.1 (Photon Systems Instruments, Brno, Czech Republic).
Determination of superoxide
Superoxide was determined with the dye MCLA (2-Metyl-6-(4-Methoxyphenyl)-3,7- Dihydroimidazol[1,2-A]pyrazin-3-One, Invitrogen). Two leaves from the 5th nodium were incubated in 1990 µl SMNS and illuminated for 30 min at 26 µmol photons m-2 s-1. The leaves were removed and 10 µl 100 µM MCLA were added to the medium to achieve a concentration of 50 nM. One molecule MCLA bound to one molecule O2•-
and generated one photon, which was detected using a luminometre (LKB WALLAC 1250 + KEITHLEY 177) and the software Oxywin 3.1. Superoxide in medium incubated without leaves and the baseline (blank value, no sample) were also measured. The relative superoxide production was defined as (value from leaves - value from medium) / (value from medium - baseline).
The leaves from the superoxide measurement were frozen in liquid nitrogen and stored at -80°C after use for pigment analyses.
Determination of H2O2 production with PF2
Intracellular hydrogen peroxide (H2O2) production was determined with a H2O2-specific fluorescent indicator based upon a boronate deprotection mechanism (Chang et al., 2004;
Lippert et al., 2011), Peroxyfluor-2 (PF2, Dickinson et al., 2010). Two leaves from the 5th nodium counted from the apex of the plant were incubated in 100 µM PF2 in 0.5 ml of SMNS for 30 min in the dark. After 30 min of destaining in 15 ml SMNS in the dark, the leaves were fixed to the measuring chamber of the FKM. The medium was exchanged after every measurement and all tubes were washed with ddH2O. The leaves were frozen afterwards in liquid nitrogen and stored at -80°C until pigment analyses. The H2O2-specific fluorescence was measured in the FKM using the filter set from AHF (Tübingen, Germany) with an excitation filter 420-500 nm (AHF F42-468), dichroic mirror 505 nm (AHF F71-302) and 520-550 nm emission filter (AHF F47-535). Flashes of blue supersaturating light were given at increasing signal integration times (20 µs up to 20 ms). For each sample, the integration time that led to the highest signal intensity without saturating the camera was chosen for quantitative analysis to avoid noise at too short exposure times, and oversaturation of the camera at too long exposure times. For each exposure time, hundreds of single pictures were taken and averaged. Background signal for each exposure time was subtracted automatically via a measurement without light.
Images of the measurement were analyzed with the FKM software and the fluorescent signal was re-calculated to one integration time according to an empirical calibration of exposure times vs. signal intensity.
Harvest
After 6 weeks of treatment, the plants were washed in SMNS and young and old tissues were separated from each other. Young tissues being 4 cm from the apex and 2 cm from the apex of each branch, old tissues 8 cm from the stem end and the rest of the branches. Remaining SMNS was removed by shaking, and the plants were frozen in acid washed tubes in liquid nitrogen. Samples were stored at -80°C until further analysis.
Determination of pigment content
Harvested material and weekly frozen leaves were lyophilized and ground with sand and a few grains of Bis-Tris (Sigma-Aldrich, St Louis, MO, USA). Pigments were extracted in 1 ml 100% acetone over night at 4°C in the dark. Spectra of pigment extracts were measured with the UV/VIS/NIR absorption spectrophotometer Lambda750 (Perkin-Elmer, Waltham, MA, USA) at a spectral bandwidth of 0.5 nm, with 0.5 nm sampling interval and from 330 to 750 nm. Pigment composition was analyzed using the Gauss-Peak-Spectra method by Küpper et al. (2000b, 2007b).
Determination of starch accumulation
The starch extraction was based on the Total Starch Assay from Megazyme (AOAC Method 996.11 and ACCC Method 76.13, Megazyme, Wicklow, Ireland). The recommended protocol including pre-dissolving by KOH was downscaled for the analyses of starch in Ceratophyllum demersum (Andresen et al., 2013b; Thomas et al., 2013).
Determination of metal accumulation
The lyophilized plant material was acid digested following the protocol by Zhao et al. (1994).
All glassware used had previously been acid washed (5% HNO3). Acids for the digestion were of supra or ultrapure quality (ROTIPURAN, Roth, Karlsruhe, Germany). Analyses were done with a graphite furnace atomic absorption spectrometer (GBC 932 AA with GF 3000, Breaside, VIC, Australia). Standard solutions for Cd, Cu, Fe and Zn were diluted from AAS Standards (TraceCERT, Sigma-Aldrich, St. Louis, MO, USA). The digested plant samples were appropriately diluted to optimal detection range with suprapure (for Cd, Cu and Fe) or ultrapure (for Zn) 1.66% HCl.
Determination of elements in the solutions
Element concentrations in the nutrient solution were determined in the barrels and in the aquaria before the plants were harvested, i.e. after one week of treatment. The respective concentrations were analyzed using an inductively coupled plasma sector field mass spectrometer ICP-sfMS (Element XR, Thermo Fisher Scientific, Waltham, MA, USA). Prior to analysis ICP-sfMS parameters were optimized every day and samples were diluted 1:20
prior to analysis. The calibration with up to 5 standards was verified using the following reference materials: SLRS-5 (River Water Reference Material for Trace Metals, NRCC), SPS- SW1 (Surface Water Level 1, Spectra Pure Standards) and SRM 1643e (Trace Elements in Water, NIST). The instrumental parameters for determination of the investigated elements are reported in Andresen et al., 2013b.
Statistics
Two-way and three-way analysis of variance (ANOVA) was done in SigmaPlot 11 (SPSS Science, USA) at the significant level of P<0.05 with Cd concentration, weeks of treatment for the weekly measured data and Cd concentration, age and light condition for the data obtained from the harvested material. The Holm-Sidak method was used for multiple comparison procedure.
Results
Growth and visible symptoms of Cd toxicity
The plants treated with 200 nM of Cd under LL conditions stopped growth directly after start of the treatments. Growth of the meristems of these plants was retarded and the meristems looked compressed. Growth decreased further until the plants had to be harvested after 4 weeks (Fig. 1). From the second week on, growth stop occurred also in the second and third highest Cd concentration (50 and 100 nM). The threshold concentration of growth inhibition and visible impairment was 20 nM Cd. Under HL conditions, some growth inhibition was also observed in the low Cd concentrations and the control due to excessive growth of epiphytic algae and cyanobacteria. The sequence of visible toxicity symptoms was slightly different for the two light conditions. Under HL conditions, the visible changes of the meristem and a reduction in the internodial growth occurred for the three highest Cd concentrations already in the first week. Between week 3 and 4, the plants treated with 20 nM Cd were also affected in plants from LL and HL conditions. All plants treated with 10 nM or less Cd did not show visible symptoms of toxicity.
Photosynthetic parameters
Cadmium treatment led to a diminished Fm under HL conditions from the first week of treatment (200 nM; Fig. 2) with subsequent reductions until 20 nM Cd in the fourth week (P<0.001). The reduction in Fm was consistent with a decreased Chl concentration with similar pattern. Under LL conditions, the effect was not as pronounced, but significant for the plants treated with 100 nM Cd.
Cadmium treatment led to a diminished photochemical efficiency of PS II as measured by Fv/Fm (Fig. 2). From the second week of treatment on, there was a distinct reduction observable: Plants treated with 50 nM, 100 nM or 200 nM Cd had decreased values, while no effect was observed below 50 nM (P<0.001). From the third week on, the plants treated with 100 nM and 200 nM Cd yielded values reduced by 50% compared to the treatment start and the control samples. The reduction in Fv/Fm was much more apparent in plants from the HL experiments compared to LL. A noticeable decrease was only observed in the macrophytes treated with 100 nM and 200 nM from week 3 on.
Photochemical fluorescence quenching displays the operating efficiency of PS II (ΦPS II, or Fv’/Fm’). The same trend as for Fv/Fm was visible for ΦPS II in HL at the beginning of the actinic light phase (i1), after acclimation in actinic light (i6), and during recovery in darkness after actinic light (r1 and r5; Fig. 3). The decrease was observed from the second week on for all plants cultivated with 50 nM or more for all 4 parameters. A different behavior was observed in the LL experiment. Photochemical quenching in actinic light was reduced in all plants of Cd treatment with 50 nM or higher from the first weeks of treatment on (Fig. 3). In the last weeks, also the plants treated with 20 nM showed inhibition. In the dark phase after actinic light (r1), a strong decrease was observed only for the plants treated with the highest Cd concentration. After approximately 200 s in darkness (r5), hardly any effect was observable, besides a reduction for the three highest concentrations (P<0.001). The other plants were able to relax after the actinic light phase.
Unlike the photochemical quenching, there was no clear trend detectable in the non- photochemical quenching (NPQ, Fig. 4) of the HL and LL plants during actinic light (i). At the end of the relaxation period (r5), however, the plants treated with 10-50 nM Cd had enhanced NPQ from the third and fourth week on, which was also observable at the spectrally resolved level (Fig.5). An enhanced quenching occurred at two positions, at approximately 690 nm and 750 nm in the plants treated with 20-100 nM towards the end of the treatment (week 5+6).
Oxygen exchange
After six weeks of treatment, oxygen exchange was measured for plants in darkness and in increasing light intensities. Net oxygen evolution increased with increasing irradiances. Cd treatment led to delayed oxygen evolution in HL plants, most prominently already from 2 nM onwards, and LL plants from 20 nM Cd (Fig. 6). Measurement of oxygen release in low irradiance led to higher amounts of O2 in plants grown under HL conditions, compared to LL grown plants. Although LL grown plants have bigger antenna systems, they have comparably fewer RC than HL grown plants (Taiz and Zeiger, 2007) and therefore less oxygen evolution.
Respiration showed no clear trend under HL conditions (measured after 6 weeks of treatment) suggesting that the respiratory electron chain in mitochondria was not so much affected by Cd treatment compared to photosynthesis. Under LL conditions, respiration after actinic light was increased in the highest measureable concentration (50 nM). Generally, oxygen uptake was higher in the plants of the HL conditions (P<0.001).
Production of reactive oxygen species
Cadmium treatment enhanced the release of superoxide (O2•-
) from the leaves into the medium (Fig. 7, upper part). The relative production of superoxide was higher and started earlier (from second week on) from those leaves of the HL experiment. Despite some noise, the threshold concentration for enhanced O2•- under HL conditions was 10 nM Cd, although significant differences were obtained between the 50 nM and all plants from 0.2 nM-10 nM.
In the leaves from LL, an increase was only observed in the three highest Cd concentrations in the beginning and towards the end of the treatment (week 5-6; Fig. 7, upper part).
Superoxide production in those three treatments was different from the control treatment (P<0.001).
Hydrogen peroxide (H2O2) was elevated in response to Cd treatment in HL plants from the second week on when treated with 50 nM Cd or higher (Fig. 7, lower part). After the fourth week of treatment, also the plants treated with 20 nM had higher relative values. In LL plants, relative H2O2 production was higher compared to HL, but was not observed below 50 nM of Cd (P≤0.002). A slight induction was observed from the second week on, but was much more pronounced after 3 weeks of treatment.
Pigment composition
Cd treatment led to decreases in most pigments (Fig. 8). Plants grown under LL conditions tended to have more light harvesting pigments (like Chl and violaxanthin), while those from HL conditions had relatively more stress-related and energy-quenching pigments (like β-carotene, or lutein). The most abundant pigment was Chl a, but the same trends occurred for total Chl, Chl a, and Chl b. Under HL conditions, the amount of total Chl decreased subsequently after the first week of treatment with 20 nM or higher, while there was no change in the plants treated with less than 20 nM. Under LL conditions, there was a drop in Chl content for all concentrations after the second week, which continued for the plants treated with 50 nM or higher, while there was a full recovery until 10 nM. The threshold concentration was 20 nM, resembling the results of Fm. The same trend was observed for neoxanthin, an energy-quenching pigment. The amounts of β-carotene-like pigments (β-carotene and zeaxanthin; which cannot be distinguished by UV-Vis spectroscopy) were affected by Cd treatment under HL conditions after the second week of treatment from 50 nM onwards. Below 50 nM, the amounts stayed rather constant until the end of the treatment.
Under LL conditions, there was a decrease like for the total Chl, without recovery, but with even more decrease in the concentrations below 10 nM.
Starch accumulation
Strach accumulation shows the long-term balance between energy production and consumption. It was higher in HL grown plants than in LL grown ones (Fig.9; P<0.001). In HL and LL grown plants, highest amounts of starch were measured in the young tissue of the plants treated with 200 nM Cd, which had to be harvested after 4 weeks of treatment in all experiments and were apparently unable to utilize the previously synthesized starch.
Metal concentration and accumulation
Cadmium accumulation in the tissues increased with increasing applied Cd concentration (Fig. 10) but under both light conditions saturation was observed for the three highest Cd concentrations (50-200 nM) with values up to 5000 ppm. There was a statistically significant interaction between Cd concentration, age and light (P=0.026). The concentrations of Cd in the barrels in the nutrient solution increased linearly with increasing concentration. In the aquaria, there was a strong increase at 20 nM Cd, suggesting that plants treated with less than
20 nM removed Cd out of the solution by uptake. The highest ratio of metal in the barrel vs.
metal in the aquarium (indicating most efficient uptake) was found at 5-10 nM Cd in the barrel.
Copper accumulation varied between 10 and 45 ppm. A decrease in Cu accumulation at increasing Cd stress was mainly observed in the young tissues from the LL treatment, with significant differences between the treatment with 100 nM Cd and the low Cd concentrations (0.2-5 nM, P≤0.001). Differences in Cu accumulation occurred due to the different light conditions (P=0.033) as well as due to age of the tissues (P=0.018). Cu concentrations were slightly lower in the barrels (~ 7.5 nM) than their nominal values should be (10 nM). Higher values in the aquaria could indicate a loss of Cu from the plants. After one week of Cd treatment, there was more Cu in the aquaria of the higher Cd concentrations (100-200 nM).
As the respective plants contained less Cu in their tissues, they either may have lost Cu from the tissue (as shown for Cu-limited C. demersum, Thomas et al., 2013a). The general bad status of the plants may have prevented Cu uptake, but this would not explain the increased (compared to the supply barrel) Cu in the aquaria.
Concentrations of iron within the tissues varied between 50 and 100 ppm, with the exception of the plant treated with 200 nM of Cd. There, higher values were determined in the old tissues (220 ppm in HL plants, 270 ppm in LL plants). Concentrations in both media from the barrels and the aquaria did not differ much due to Cd treatment.
The measurements of Zn in the tissues were rather noisy and no differences were observed due to Cd treatment. This was also observable in the nutrient solution. Although nominal 100 nM Zn should be present in the nutrient solution, in average only 70 nM were observed. Increasing values in the aquaria at the higher Cd treatments may be due to reduced Zn uptake caused by blocking of the transporters by Cd.
As the plants were not washed in EDTA (or a similar chelating agent), the values reported here include binding of the metals in the cell walls. However, as this is true for all samples, the trends in intracellular metal accumulation are nevertheless observable.
Discussion
Time sequence, concentration dependency and causal relationship of Cd toxicity
The first physiological process inhibited by Cd treatment was photosynthesis (Fig. 2, Fig. 8).
Under both HL and LL conditions, the plants treated with the highest Cd concentration
showed reduced values for Fm, total Chl and Fv/Fm, ΦPS II i6 and r5 already after the first week of treatment with high Cd concentrations. Apparently, PS II photochemistry (Fv/Fm) as well as operating efficiency (Φ) and further on the linear electron transport and ultimately CO2
assimilation were affected. This suggests an impairment of photosynthesis and electron transport. A major inhibition side of heavy metals is the replacement of the Mg2+ ion in Chl.
The formation of heavy-metal substituted [hms]-Chl was observed for many photosynthetic organisms (reviewed by Küpper et al., 2006). Those [hms]-Chls are unsuitable for photosynthesis for many reasons (see Küpper et al., 2006 for details). The substitution of Mg2+ in the Chls was shown to be different depending on the light conditions (Küpper et al., 1998). Under LL conditions, plants have bigger antenna systems (see Fig. 8; P < 0.001 of harvested material) with more light harvesting complexes (LHCs) per reaction center (RC) (Taiz and Zeiger, 2007) and the substitution will take place mainly in the LHCs. Indeed, our results confirm that under LL, the fluorescence data were less reduced by Cd treatment than under HL conditions, where the insertion of Cd in the PS II RC (most likely into the pheophytin, Küpper et al., 2002) is more likely (Fig. 2 and 3). By this, the core is destroyed and the whole photosystem is lost. Even if only a small percentage of RC-Chls get substituted by another divalent ion, photosynthesis can be strongly inhibited (Küpper et al., 2006).
Furthermore, Cd-Chl is, unlike Cu-Chl, very unstable, and will be degraded. The decrease in Chl content occurred under both light conditions in this study at concentrations of 20 nM or above, already after one week of Cd treatment (Fig. 8) suggesting the degradation of dysfunctional Chl.
Respiration was measured only at the end of the treatment, but no clear inhibition due to Cd was observed (Fig. 6), suggesting that the mitochondrial electron transport chain is less susceptible than the plastidal one. The complexes of the respiratory chain contain almost exclusively Fe atoms (as heme or Fe-S-cluster), which will not be substituted by Cd atoms (Moulis, 2010). That Cd has an effect on isolated mitochondria was shown already in 1973:
mitochondria from corn shoots yielded lower oxidation rates (Miller et al., 1973). But the analysis of actual presence of Cd within mitochondria in whole plant samples is still lacking.
Van Belleghem et al. (2007) used energy-dispersive X-ray microanalysis (EDXMA) on Cd treated (1, 5 and 50 µM) A. thaliana leaves and roots and detected Cd as fine deposits in the vacuole and as large granular deposits in the cytoplasma, but not in any organelle. But every technique where embedding in a resin or dehydration is involved (like freeze substitution and embedding in Spurr’s epoxy resin) the artefactual redistribution of elements cannot be prevented. The detection of ions in specific organelles is therefore impossible in such
