3.1 A systematic characterisation of C. reinhardtii 28 strains
With the aim to increase the selectivity of algae-based biosensors we propose a systematic study of a collection of 28 strains of the green alga C. reinhardtii to exploit as an array of biocomponent with different affinity towards diverse herbicide classes. The intention is to provide a multiplexing analysis of target analytes of agro-environmental interest by the evaluation of inhibition studies with solutions and mixtures of herbicides.
The choice of the bioreceptor during biosensor design is an essential phase, which deserves the right assessments in terms of sensitivity, repeatability, and stability among other analytical parameters. To meet these demands, many inspirations have been originated by docking and molecular genetics approaches, to obtain very sensitive and selective bioreceptors. In the last decades, more than 50 different C. reinhardtii algae strains have been collected in our laboratories and used for different purposes (e.g. nutraceutical, space research, and biosensing) (Giardi et al., 2013). This handbook on biosensor design started from the selection of 28 different strains of C. reinhardtii from this collection, obtained in previous studies by diverse techniques as cloning, mutagenesis, and biolistic transformation. Among them, CC125 is the wild type strain exploited as control (Elisabeth et al., 1987). This corresponds to the basic "137c" wild type (isolated in 1945 near Amherst MA, USA, by G.M. Smith) and is presumably equivalent to strain 11/32c (Culture Centre of Algae and Protozoa, location, country). CC125 is the background strain for many mutations, including CC503 cw92 mt+, which has been the source of DNA, used for genomic sequencing. CC125 has been kindly provided by the Chlamydomonas Collection, Duke University to UTEX in 1980 (UTEX 2244).
CW15 is a cell-wall mutant with a particular defect in cell wall production (Davies et al., 1971; Loppes et al., 1975).
The intron less (IL) strain has been produced from wild-type cells 11/32b using an efficient chloroplast transformation protocol. In particular, the highly conserved psbA gene has been manipulated, and a transformant with no introns in the psbA gene has been generated starting from wild-type cells 11/32b (Sammlung von Algenkulturen, Grttingen, FRG). Starting from this strain, the deletion mutant FuD7 and the Del mutant have been obtained in previous studies as well as a collection of different mutants (Johanningmeier and Heiss, 1993). For this reason, IL is considered as a reference strain for the 25 D1 mutants selected for this protocol and reported in Table 1.
The selected 28 strains were physiologically characterised, by evaluating cell culture growth, pigment content, photosynthetic activity, oxygen evolution, and herbicide detection capability, with the aim to exploit them as bioreceptors for the realisation of a dual electro-optical transduction algal biosensor.
3.1.1 Cell culture growth
The growth of selected Chlamydomonas strains was characterized by measuring the optical density of the culture over a 165-hour period at a 750 nm wavelength. Figure 2 reports the time courses of all the 28-strain culture growth, showing the difference among the grow rate of D1 mutants, the reference strain IL, the wild type control CC125, and the mutant cw15. As shown, during the analysed period, absorbance values were found similar for most D1 mutants, and slightly higher for A153S, A250V, A251C, probably due to the particular amino acidic substitution, as well as for the wild type CC125 and cw15 mutant. Values are the average of 3 independent experiments (±SE, n=3).
3.1.2 Pigment content, Photosynthetic activity, and Oxygen evolution
The pigment content of selected Chlamydomonas strains was analysed by following the chlorophyll a and b accumulation of the cultures over a 165-hour period under continuous light (50 µmol photons m-2 s-1), agitation (150 rpm) and 25 °C temperature, obtained by 80% acetone extraction and spectrophotometrically measured as absorbance at a 652 nm wavelength. As depicted in Figure 3A, the chlorophyll content in the wild type CC125 was, in the majority of cases, more than three times as high as in most mutants, excluding IL and A250R. The photochemical efficiency of C. reinhardtii strains was evaluated by the fluorescence analysis of the chlorophyll a described by the OJIP or Kautsky profile vs time, by which it is possible to calculate the characteristic photosynthetic parameters FV/FM (maximum fluorescence yield of PSII) and 1-VJ (relative variable fluorescence at 2 ms). The values of these fluorescence parameters were reported in Figure 3B, calculated by each Kautsky curve and plotted as a function of time, to show the photosynthetic efficiency of the different strains in physiological conditions. These results evidence that some strains (e.g. CC125, IL, A250R, A250V, CW15, S264K) exhibit a higher photosynthetic activity in terms of maximum fluorescence yield in comparison with other strains (e.g. A153S, A250L, L159M, S212C among others) which show a reduction of this parameter. On the contrary, some strains, with a sufficient maximum fluorescence yield, show decreased values of the variable fluorescence parameter, as for example the strain S264K. In general, the variations of the photosynthetic parameters could be ascribed to a rearrangement of the photosynthetic apparatus due to the induced mutations in the D1 protein, as well as to a change in electron transfer rate along with the photosynthetic chain, in particular between the first electron acceptor QA and the second one QB. As a result, the evident variation in shape of the OJIP curve of these mutants with respect to the control strain is directly attributable to the accumulation of reduced QA and to the marked slowing of electron flow.
Oxygen evolution for the 28 C. reinhardtii strains was amperometrically measured on liquid cell cultures exploiting carbon black (CB) nanomodified screen-printed electrodes (SPEs) as a valid alternative to the conventional Clark electrode. CB nanoparticles, indeed, show their potential to sense oxygen reduction thanks to their ability to electrocatalyse the reduction of oxygen at lower potential than unmodified SPEs, with increase of sensitivity due to nanodimensions onion-like carbon structure, high number of defect sites and higher content in O atoms (Mazzaracchio et al., 2019; Arduini et al., 2010; Arduini et al., 2011; Arduini et al., 2020; Talarico et al., 2015). In this case, CB is able to measure oxygen produced within the photosynthetic process when algae are under illumination (350 µmol photons m-2 s-1). The rate of oxygen evolution was measured for each strain at an applied potential of -0.6 V after a period of 10 min of dark adaptation. As shown in Figure 3C, CC125, A250L and F255N strains exhibited the highest current signals up to 1.5 µA.
3.1.3 Herbicide detection capability
The response of the selected strains to triazine and urea type herbicides was studied by chlorophyll a fluorescence induction curves (OJIP or Kautsky curve). The relative variable fluorescence at 2 ms at the characteristic point J(VJ) gives a measure of the relative amount of the reduced QA. Therefore, 1-VJ represents the fraction of oxidised QA or the efficiency by which the electron is transferred from plastoquinone QA to QB. In the presence of photosynthetic herbicides, the QB binding site is permanently occupied by the inhibitor and electron transport does not extend beyond QA. The accumulation of reduced forms of QA is proportional to the herbicide concentration and can be directly evaluated by the increase of the variable fluorescence level at point J, making the parameter 1-VJ one of the most sensitive for herbicide detection. Using this approach, we evaluate the sensitivity of the selected strains to atrazine, terbuthylazine, and diuron as case study herbicides. Figure s1 (Supplementary Materials) reports the fluorescence response of the 28 Chlamydomonas strains towards the 3 herbicides as Kautsky curves as well as calibration curves expressed as 1-VJvs herbicide concentration. Moreover, with the aim to highlight the different sensitivity of all the strains, their response to a fixed herbicide concentration of 100 nM is reported in Figure 4, which underline how each strain demonstrates different inhibition of the photosynthetic activity in the presence of the herbicides.
Among other, L159V, L159M, L200I, I163N, I163S, I163T, P162S, S212C, S268C strains and the double mutant S209A/S212C showed the highest sensitivity to atrazine with a ~ 50 % of photosynthetic activity inhibition. While L159V presented the highest sensitivity to atrazine, CC125, IL, and S268C showed the highest sensitivity to terbuthylazine (~ 50 % inhibition). Whereas, S264K strain is resistant to triazine while very sensitive to urea herbicides. In addition, most strains are highly susceptible to diuron exposure, with an inhibition from 60 % (e.g. for CC125, IL, A250L, A250R, and A251C) to 80 % (for L159M, L200I, P162S, S268C). Notably, F255N strain showed high resistance to both herbicide classes.
The inhibition constant IC20 (herbicide concentration required for 20% inhibition of the parameter 1-VJ) and detection limits were also calculated (Table 2) on the basis of 99 % confidence interval, which, assuming the normal distribution of the data, corresponds to 2.6 × standard error of the measurements (σ), exploiting the modified relationship for the Langmuir absorption isotherm, LOD = 2.6×σ× IC20/(100-2.6×σ) (Koblížeket al., 2002).
3.2 C. reinhardtii immobilisation on paper-based SPEs
Once assessed the physiological features of the selected strains, further experiments for the development of the dual electro-optical biosensor have been accomplished exploiting C. reinhardtii CC125 strain as a case study.
A limiting step in the development of whole cell biosensors is the immobilization of the biomaterial in a matrix that prevents leaching, without reducing the stability and activity of the cells. Several methods are currently available for cell immobilisation, with entrapment as the most frequently used. With the scope to enhance the yield in the production of working algae-sensors as well as to achieve a higher repeatability in algae-sensor preparation, algal whole cells were immobilised on paper-based screen-printed electrodes modified with carbon black nanomaterials (paper-SPEs). The rationale behind the choice of using electrodes printed on paper consists of capturing algal whole cells into a 3D matrix composed of a natural polymer, thus providing a comfortable environmental for cells accommodation guaranteeing a better algae survival as well as an effective diffusion across the matrix of the target analyte.
To this aim, whole cells of C. reinhardtii CC125 were immobilised on the working electrode surface of paper-based screen-printed electrodes nanomodified with carbon black (paper-SPEs). To preserve the algae photosynthetic functionality, cells were entrapped in a calcium/alginate matrix. Thus, 1.2 × 107 cells of CC125 are suspended in 50 mL of 1% sodium alginate and 100 mL of 50 mM Tricine, 20 mM CaCl2, 5 mM MgCl2, 70 mM sucrose pH 7.2. Then, 5 mL of this suspension were drop cast on the working electrode of the SPE and then cross-linked using 5 mL of 200 mM CaCl2 in the same buffer, obtaining a final number of cells of 4 × 105. This calcium alginate matrix can provide a biocompatible environment for algae entrapment, without collateral effect on their metabolism; at the same time, its porous network guarantees a good diffusion of the target analyte. The algae/pCB-SPEs were thus stored in 200 mM CaCl2 in 50 mM Tricine, 20 mM CaCl2, 5 mM MgCl2, 70 mM sucrose pH 7.2 under continuous light at 50 μmol photons m-2 s-1 for successive measurements. Each algae/paper-SPE was then accommodated into the measurement cells of the electro-optical transducer for the electrochemical analysis. SEM analysis and storage stability were performed to characterise the implemented algae-SPE sensors in depth (Figure 5A and 5B, respectively).
3.3 Study on different (nano)modified SPEs and working stability
Different screen-printed electrodes modified with diverse (nano)materials were exploited to evaluate the best material for the analysis of the current signals coming from the algae oxygen evolution, including carbon black, carbon nanotubes, gold, gold nanoparticles, reduced graphene oxide, and poly(3,4-ethylenedioxythiophene) (PEDOT). Results reported in Figure 6A highlight how carbon black represents the best choice, as this nanomaterial is able to enhance the electron transfer between algae evolved oxygen and the working electrodes, thus amplifying the current signals, especially in comparison with other (nano)materials.
Once selected the best nanomodified SPE, working stability of CC125 C. reinhardtii in solution and immobilised on paper-based SPEs modified with carbon black (pCB-SPEs) was evaluated by both amperometric and fluorescence analysis, run for many hours at room temperature, under repeated cycles of 10 min dark and 30 s light (350 μmol photons m-2 s-1) in a measurement volume of 200 mL of 50 mM Tricine, 20 mM CaCl2, 5 mM MgCl2, 50 mM NaCl, 70 mM sucrose pH 7.2. Figure 6B shows 100 % of light-induced oxygen evolution activity up to ca. 8-10 hours, with a progressive decrease of signal later on. This indicates not only a very good operational stability but also a high intra-electrode repeatability with RSD of 1.1% (n=12). However, there is a clear difference between the signal stability of algae in solution and immobilised, evidencing that the immobilisation confers a higher robustness of the sensor. Figure 6C reports the Kautsky profiles of algae in solution as well as immobilised, underlining that the algae photosynthetic activity slightly suffers from the immobilisation while remaining constant during the 8-10 hours of the stability test.
3.4 Reversibility, matrix effect, and interferents
To gain further insights on the evaluation of the diverse analytical parameters, tests were performed to evaluate the nature of herbicide inhibition by both amperometric (Figure 7A) and fluorescence analysis (Figure 7D), exploiting atrazine as case study. These tests consist in washing the algae/pCB-SPE with 50 mM Tricine, 20 mM CaCl2, 5 mM MgCl2, 50 mM NaCl, 70 mM sucrose pH 7.2 after the measurement of the inhibition at an atrazine concentration of 6.6 mM, for the electrochemical analysis, and 100 nM, for the optical one, to record both the recovered current and fluorescence signals. As reported in Figure 7A/D, a complete recovery of both algae oxygen evolution and photosynthetic activity occurred.
To investigate the suitability of the proposed biosensor in real samples, matrix effect and recovery studies were also provided. To estimate matrix effect by the electrochemical point of view, algae/pCB-SPEs were incubated in 200 mL of 2×buffer (100 mM Tricine, 20 mM CaCl2, 5 mM MgCl2, 50 mM NaCl, 70 mM sucrose pH 7.2) diluted 1:2 (v:v) in surface water (Sebou River, Morocco) and fortified with atrazine in a concentration range from 0.1 to 41.6 μM. Calibration curves were obtained for both standard solutions and real samples, described by the equations y = 0.98(±0.01) - 0.06 (±0.003) x (R2 = 0.9952) and y = 0.98 (±0.01) - 0.09 (±0.005) x (R2 = 0.9945), respectively (Figure 7B). The biosensor response was in the linear range from 0.1 to 6.6 μM atrazine and showed a detection limit of 2 nM. The ratio between slopes of the calibration curves obtained in standard solutions and real samples was equal to 0.72, indicating a moderate dependence from surface water matrix (28 %). The calibration curve obtained for the electrochemical algal biosensor challenged in real samples was further used to calculate the recovery values of the surface water samples. Recovery values of 106.6 ±10% and 96 ± 8% were obtained for 3 and 5 mM of atrazine, respectively, highlighting a satisfactory capability to detect atrazine also in surface water.
Regarding the fluorescence analysis, tap water was exploited as real environmental matrix fortified with atrazine in a nanomolar concentration range. Calibration curves were obtained for both standard solutions and real samples, described by the equations y = 0.516 (±0.005) - 0.0012 (±0.00006) x (R2 = 0.9854) and y = 0.518 (±0.003) - 0.0013 (±0.00003) x (R2 = 0.9963), respectively (Figure 7E). The biosensor response was in the linear range from 10 to 200 nM atrazine and showed a detection limit of 5 pM. The slopes of the calibration curves obtained in standard solutions and real samples were very similar with a ratio of 0.92, indicating a slight dependence from the tap water matrix. The calibration curve obtained for the optical transduction challenged in real samples was further used to calculate the recovery values of the tap water samples. A recovery value of 96±5 % was obtained for 75 nM of atrazine, highlighting a satisfactory capability to detect atrazine also in tap water.
It is crucial to point out the observation of the different detection limits obtained for the two detection methods exploited, highlighting that the optical transduction provides atrazine analysis in the nanomolar range while the electrochemical one in the micromolar range.
With the aim to challenge the capability of the proposed algal biosensor in real environmental water samples, some chemicals including metals, pesticides, and phenolic compounds were tested as interferences at legal limits established by the EU Directive 2013/39/EU for surface water (where present). In detail, algae/pCB-SPEs were incubated in 200 mL of 50 mM Tricine, 20 mM CaCl2, 5 mM MgCl2, 50 mM NaCl, 70 mM sucrose pH 7.2 fortified with standard solutions of 100 ppb arsenic, 20 ppb copper, 5 ppb cadmium, 10 ppb lead, 10 ppb bisphenol A, 1 ppb paraoxon, and 6.6 mM atrazine, terbuthylazine, and diuron, as well as with a solution spiked with all the above listed compounds, to evaluate the synergistic effect of different chemicals in a mixture. Results reported in Figure 7C and 7F highlight that the interfering species did not affect the analysis of atrazine at the exploited concentrations in both electrochemical and fluorescence analysis.