The common wild type strain CC-124, also known as 137c, was chosen for this study and obtained from the Chlamydomonas Resource Center¹. Growth conditions varied throughout trials in terms of media, vessel type (flask or microplate), microplate shaker speed (flasks were grown at 100 rpm), and photoperiod as indicated. Cultures were maintained at room temperature (≈22°C) with an overhead fluorescent light source (250 μmol photons/m²/s) with Tris-Acetate-Phosphate (TAP) media or Tris minimal media (reduced nitrogen) (Gorman and Levine, 1965). Liquid cultures were inoculated from pre-grown plates, and grown for 48 h.

Aliquots from these cultures were diluted to ≈2.5 × 10⁵ cells/mL at the start of each experiment. C. reinhardtii was grown in either 125 mL flasks or 96 well microplates in either 1/5 (25 mL) or 2/3 (200 μL) total liquid volumes, respectively, to ensure aeration. An improved Neubauer hemocytometer and microscope (200–400× magnification) were used to measure cell concentrations manually. Flask and microplate cultures were grown for 72 h under day:night (16:8 h) and continuous photoperiods examining different microplate shaking speeds (100, 150, and 300 rpm). Each sample type had three biological replicates. Cell concentration was measured with a hemocytometer at each timepoint (0, 3, 6, 9, 12, 24, 36, 48, and 72 h) in triplicate. Growth curves were plotted from these results and were compared to the 100 rpm flask with the R statsmod compareGrowthCurves function with a threshold of p < 0.05. Mixing between wells as well as evaporative loss was prevented by using clear plastic plate covers that were sterilized with 70% ethanol.

The absorbance spectrum of C. reinhardtii cultures (TAP + 0.5% DMSO) was examined at 1 nm intervals from 300 to 800 nm using an i3 Spectramax plate reader. Cell concentrations were measured in triplicate with a hemocytometer, and OD₅₅₀ was also recorded for each individual sample. Manual measurements were correlated to OD₅₅₀ measurements with a third order polynomial, the raw OD₅₅₀ values were converted through their respective polynomials to generate predicted values, and the predictions were then compared to hemocytometer measurements with regression analysis.

Cultures were grown in either flask or microplates (100 and 300 rpm, respectively) with TAP, TAP-DMSO (0, 0.5, 1, and 2% v/v), or Tris-minimal media (low nitrogen) (Gorman and Levine, 1965). Each sample type had four biological replicates. Cell concentrations were manually measured with a hemocytometer at 0, 24, and 48 h, and then analyzed with either a Welch’s ANOVA and Dunnett’s t-test post hoc comparison at each timepoint to the flask TAP control, or an independent t-test at each timepoint between vessels of the same media type.

Fluorescein diacetate (FDA λ_ex: 493 nm, λ_em: 523 nm) was added at a final concentration of 2.4 μM (1 μg/mL) to C. reinhardtii cultures (TAP + 0.5% DMSO). Samples were incubated in the dark at room temperature for 30 min before measuring fluorescence. Heat treating the samples for ≈45 min at 90°C was used to establish cell death and act as a negative control for viability (Vavilala et al., 2016). A modified version of FDA, 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCFDA), was used to measure reactive oxygen species (ROS) at a final concentration of 100 μM (48.73 μg/mL). Incubation times and measurement settings were identical to FDA.

Two Promega (Madison, WI, United States) viability kits were also evaluated – CellTox Green Cytotoxicity Assay (membrane integrity) and RealTime-Glo MT Cell Viability Assay (reducing potential). These kits were multiplexed, testing both assays simultaneously in each sample, in accordance to manufacturer instructions as an endpoint assay following heat treatment. All viability assays were performed on an i3 Spectramax plate reader. Cell concentration and percent viability were recorded, and a Welch’s ANOVA was performed for each at 48 h.

Photopigment concentrations were measured with the standard Lichtenthaler 80% acetone extraction method – cell cultures were pelleted through centrifugation (8 min, 10,000 × g), resuspended in an equal volume of 80% acetone, vortexed (30 s), centrifuged under the same conditions again, and the resultant supernatant was measured spectrophotometrically at 470, 647, 663, and 750 nm (Lichtenthaler, 1987). These wavelength measurements were blanked by subtracting A₇₅₀ from each other value before calculating pigment concentrations using the following equations: ChlA (μg/ml) = 12.25(A₆₆₃)–2.79(A₆₄₇), ChlB (μg/ml) = 21.5(A₆₄₇)–5.1(A₆₆₃), Total carotenoid (μg/ml) = [1000(A₄₇₀)–1.82(ChlA)–85.02(ChlB)]/198. For in vivo measurements, cell cultures were directly measured at 470, 650, 680, and 750 nm with A₇₅₀ used again as a blank. The extraction values were correlated to the in vivo values for each pigment with third order polynomials, and regression analysis was also performed between the predicted and actual measures.

Silver nanoparticles were synthesized by a procedure modified from those described by the California NanoSystems Institute². In brief, silver nanoparticles were synthesized by heating a solution consisting of equal volumes silver nitrate (0.0047 M) and sodium citrate (0.20 M) at 80°C for 30 min.

Cultures of C. reinhardtii CC-124 were grown in both flasks and microplates to determine the effect, if any, exerted by container type. Cultures were grown for 72 h with either a 16:8 h day:night cycle (Figure 1A) or a continuous photoperiod (Figure 1B). The 16:8 photoperiod reflects natural lighting conditions, while the continuous photoperiod is used to maximize growth for industrial applications. Sterile plate-seals were used to prevent evaporation. Considering the importance of proper aeration to C. reinhardtii growth, microplate cultures were grown at 300 rpm initially to better mimic flask aeration (100 rpm). Cell concentrations were manually measured, by hemocytometer, in triplicate at 0, 3, 6, 9, 12, 24, 36, 48, and 72 h intervals. A growth curve for each photoperiod (16:8 day:night or continuous), shaker speed (in rpm), and vessel type (flask or plate) is shown in Figure 1A (16:8) and B (continuous). The 300 rpm samples actually exceeded the growth of C. reinhardtii in flasks under both lighting conditions. Reducing the shaking speed to 150 rpm resulted in microplate cultures which mimicked growth in control flasks under similar lighting conditions (Figures 1A,B).

Manually determining C. reinhardtii cell concentration is time consuming and subject to sampling and human errors; automating these measurements would reduce these negative factors and is necessary for making this organism viable for any sort of high-throughput, real-time assays. To correlate manual cell concentration counts to absorbance we first needed to identify an appropriate wavelength, one generally free of existing signal. The full absorbance spectrum of C. reinhardtii cultures in 1 nm increments noted a relatively signal free region around ≈550 nm between the chlorophyll peaks (Figure 2A). Using this wavelength, and a known series of culture dilutions, we developed a correlation curve to predict cell density with an R² ≈ 0.94 (Figure 2B). This assay accurately predicts cell concentrations from 2.8 × 10⁵ cells/mL and higher with an average absolute value of percent deviation from known at 16.16% ± 1.38 (SE). This is an acceptable range for preliminary plate-based screens to provide leads for further testing.

Nutrient availability is an important consideration in C. reinhardtii growth, motility, and for biotechnology applications, such as biofuel production. TAP is a relatively, nutrient rich media for the growth of C. reinhardtii, so we next wanted to test growth under reduced nutrient conditions. Tris-minimal (nitrate free) media has previously been shown to reduce growth in flasks (Dean et al., 2010). Indeed, at 24 h in Tris-minimal media, flask growth was reduced by approximately 50% relative to TAP controls, while microplate growth was not significantly altered (Figures 3A,B). However, at 48 h both flask and microplate growth in Tris-minimal media were comparably reduced relative to TAP controls. These findings were comparable across both the continuous as well as the 16:8 day:night cycle.

Dimethylsulfoxide (DMSO), is a common carrier solvent for small molecule screening in biological assays, and has been previously evaluated in C. reinhardtii (Alfred et al., 2012). However, DMSO becomes toxic to cells at higher concentrations, and a dosing range which would not alter the growth of Cr needed to be established (Alfred et al., 2012). Microplate cultures were therefore evaluated for growth at 0, 0.5, 1, and 2% DMSO (v/v) (Figure 4). Increasing DMSO concentrations lead to a decrease in cell concentrations, most notably at 24 h (50–80%), and to a lesser extent at 48 h. 0.5% DMSO did not appreciably change compared to the untreated control. We conclude that 0.5% DMSO and not 2% is an acceptable concentration range for studies with small molecules at 24 and 48 h timeframes as this ensures a similar growth ‘trajectory.’

While growth is an easy measure of toxicity, it provides little insight into stress related effects which may appear at lower concentrations. Numerous viability assays for unicellular organisms are available, but must be evaluated on a case by case basis. For example, the hydroxyproline rich cell wall of C. reinhardtii may act as a barrier to several of the common cell viability kits used with microplate-based systems (Azencott et al., 2007). FDA, a fluorescence-based viability indicator used previously in C. reinhardtii was chosen as a standard for comparison with other viability markers (Amano et al., 2003). Neutral FDA is able to diffuse across cell membranes and is subsequently cleaved by cytoplasmic esterases to yield anionic fluorescein (λ_ex = 475, λ_em = 535) which is typically restricted to the cell. As shown in Figure 5, significant fluorescence was observed in control wells with only minimal background fluorescence, while heat treated samples showed <5% of the fluorescence of the controls, consistent with the loss of esterase activity.

In order to expand the toolbox available for viability assays we included three additional approaches to assessing cell viability and/or stress: H₂DCF-DA (2′,7′-dichlorodihydrofluorescein diacetate), as well as Promega’s CellTox Green Cytotoxicity and RealTime-Glo MT Cell Viability assays. Like FDA, H₂DCF-DA diffuses across the membrane and reacts with cellular esterases, the subsequent H₂DCF is then subject to oxidation by ROS to yield fluorescent DCF (2′,7′-dichlorofluorescein). As shown in Figure 5A, viable cells had a positive fluorescent signal, while heat-killed cells showed substantially less fluorescence (<5% of treated controls), consistent with these cells being dead, and therefore unable to engage in esterase activity. The addition of 200 μM CuSO₄, a known stimulator of ROS production, served as a positive control for the assay (Stoiber et al., 2013).

The RealTime-Glo assay measures the reducing potential of cells through a luminescent substrate. Conversely, the CellTox Green Cytotoxicity assay measures viability by using a membrane impermeable DNA binding indicator. Increased fluorescence in this assay arises from the fluorophore intercalating with DNA which can only occur when the plasma membrane integrity is compromised, a sign of cell stress or potential viability loss. Both of these assays were viable in C. reinhardtii, showing increased (CellTox) or decreased (RealTime-Glo) signals, respectively, in heat treated samples relative to untreated controls (Figures 5A,B). We hypothesized that the reduced growth observed in 2% DMSO may well be reflected in stress/viability assays, specifically cell permeability given this chemical’s use as a carrier solvent often for lipophilic substances. Indeed 2% DMSO treatments resulted in an increase in the CellTox Green fluorescence, consistent with this hypothesis but well below the heat treated threshold (Figure 5B). 0.5% DMSO, however, had no apparent effect on membrane permeability by this method, consistent with this concentration being less disruptive.

The photopigments, chlorophylls and carotenoids, provide valuable insight into photosynthetic potential, metabolic flux, and cell stress (Lichtenthaler and Rinderle, 1988; Boyle and Morgan, 2009; da Silva, 2016). Our plate based assay has the potential to measure relative photopigment concentrations in a non-destructive in vivo manner. Such non-destructive approaches would be beneficial for probing environmental, chemical, or biological effectors of C. reinhardtii. To develop a non-destructive assay for measuring photopigment concentrations we began by evaluating changes in the in vitro and in vivo absorbance maxima for each pigment. An absorbance spectrum of intact cells indicated a red (right) shift of the chlorophyll a (A₆₆₃→A₆₈₀) and b (A₆₄₇→A₆₅₀) maximas in vivo, relative to extracted pigments in 80% acetone (Lichtenthaler, 1987). Such bathochromic shifts are well documented in the literature and are generally accepted to arise from differences in the chemical environments of the pigments between in vivo and cell lysates (Morosinotto et al., 2003; Wang et al., 2008). These red shifted values (680 and 650 nm, respectively) were used in place of the customary in vitro (663 and 647 nm, respectively) values to generate representative pigment amounts (Figure 6). No significant red-shift was observed for carotenoids which were visible at ≈470 nm.

Representative chlorophyll a, b, and total carotenoid values were correlated to known values, and the predicted values corresponded to the actual with R² values of 0.999 for each. Predicted pigment abundance was normalized to extraction values, and samples above 2.18 × 10⁶ cells/mL closely follow known values, while below this threshold the equation loses predictive power. We thus concluded that this in vivo assay produces accurate predictions of photopigment abundance with a lower limit of quantification at cell concentrations of 2.5 × 10⁶ cells/mL, and an average absolute value of percent deviation from known at 2.52% ± 0.68 (SE).

Having completed the development of our plate assay, we wanted to validate its utility through a toxicity screen. Specifically, we wanted to distinguish differences in the toxicity of free silver cations and silver nanoparticles. Nanoscale materials possess unique properties due to their small size, large surface area-to-volume ratio, and quantum effects. As a result nanoparticles may be more or less toxic than an uncomplexed (free) version of the same element and distinguishing between these relative toxicities is important. Engineered nanoparticles are already found in over 2000 consumer goods and this number is to expand significantly in the next decade. It is therefore important to develop an affordable assay for screening nanoparticle toxicity and our C. reinhardtii screen is ideal for this purpose. Furthermore, this organism has previously been utilized for analyzing the toxicity of metal nanoparticles making this an excellent reference study for this purpose (Leclerc and Wilkinson, 2014; Navarro et al., 2015).

In order to evaluate and compare toxicities, we co-incubated C. reinhardtii cultures with either free silver cations or synthesized silver nanoparticles (≈40 nm) for 48 h at four different concentrations (0.1, 1, 10, and 50 ppm). After 2 days of growth we evaluated all our samples for changes in growth, photopigment, and ROS production (H₂DCFDA) relative to untreated controls. Silver nanoparticles were noticeably more toxic than free silver, inhibiting growth by 25% at even 1 ppm (Figure 7). Both treatments increased ROS and carotenoid production at concentrations as low as 0.1 ppm, before any negative effects on growth were observed. However, this was followed by a significant decline in overall photopigment production as toxicant concentration increased. Despite these similarities some differences in mechanism appear to exist, as 50 ppm of silver nanoparticles clearly disrupted cell membrane integrity with the CellTox assay (sixfold increase in fluorescence, ≈3000 RFUs vs. 18000 RFUs, untreated vs. treated) while ‘free’ silver had no significant effect on fluorescence relative to untreated controls.