The Effects of Environmental Stressors on Aggregation in Parachlorella kessleri

The induction of biofilm formation has been explored as a means of harvesting microalgae for bioprocessing applications. Environmental stressors have been implicated in the induction of biofilm formation; however, it is unclear whether all stressors, or a select few, are responsible. This study aimed to in-vestigate the effects of three stressors, nitrogen depletion, reduced or elon-gated day lengths, and increased culture turbulence on biofilm formation of Parachlorella kessleri. We also examined corresponding effects on growth and production of reactive oxygen species. Turbulence induced the greatest response in which a significant decrease in growth plus an increase in superoxide production and flocculation efficiency were seen for the 300-rpm treatment. For varying day lengths, stress response was not observed, howev-er, a significant increase in EPS secretion was measured in both shorter and longer days. Nitrogen depletion induced a low-level stress response, in which superoxide production increased for the highest concentrations, while growth was not impacted. In contrast to previous studies on nitrogen depletion, a significant increase in EPS secretion was not observed. Results indicate that stress response varies according to type and magnitude. EPS production and thus biofilm formation are not linked to the stress indicators investigated. P. kessleri uses small quantities of EPS to contribute to cell stickiness, but not necessarily to the full formation of a biofilm; however, cell stickiness served as a mechanism for substrate adherence and cellular aggregation none the less.


Introduction
Newly available environmental substrates are colonized through a successional How to cite this paper: Schriber, C.L. and Venable, M.E. (2019) The Effects of Environmental Stressors on Aggregation in Parachlorella kessleri. Journal of Environmental Protection, 10, 1657-1676. Nutrient deprivation often leads to a surplus of carbon within the cell, and thus increasing the C:N ratio. CO 2 -fixation continues unchanged but the carbon cannot be used for protein synthesis and is funneled into carbohydrates [6]. The excess carbohydrate from carbon overflow is stored intracellularly until it exceeds that which the cell can store, and is secreted extracellularly as EPS [22]. Glyoxylate simulates nitrogen deprivation through the acceleration of photosynthesis and the inhibition of photorespiration, without augmenting N-metabolism [20] [23]. Yang et al. [6] showed that glyoxylate increased extracellular polysaccharide content and an increase in cellular aggregation and size in Chlorella sorokiniana (pyrenoidosa).
Microalgal polysaccharide production may be linked to a circadian clock [24].
Cell stickiness of Chlamydomonas reinhardtii has been shown to increase at night [25]. Day length changes affect photosynthetic processes and carbon metabolism and intra-and extracellular polysaccharides. Cellular responses to a variation in light intensity or duration show overlap [26].
Grobbelaar [27] showed that turbulence elicits beneficial aspects of mixing, nutrient redistribution, access to higher quality light, and the removal of waste as well as detrimental effects of shear stress, through mechanical disruption of the cellular membrane and/or wall [28] [29]. Mitsuhashi et al. [30] found that photosynthesis by Parachlorella kessleri is also inhibited by turbulence through hydrodynamic shear stress, a form of disturbance associated with physiological disruption as opposed to mechanical. As is often found for microalgae, the response to turbulence stress is species-specific, where green microalgae have been found to be more tolerant [9] [28].
Decreases in growth in response to turbulence have been observed in Parachlorella kessleri and Scenedesmus quadricauda [31]. Rodríguez et al. [32] showed that turbulence stress led to an increase in lipid peroxidation. Turbulence-induced biofilm and EPS viscoelastic properties have been linked to protection from shear stress [5] [21] [33].
The research reported here investigated whether simulated nutrient, photoperiod or turbulence stresses induce aggregation in Parachlorella kessleri by measuring the flocculation efficiency and polysaccharide content. This investigation also determined if common stress responses, such as growth inhibition and ROS production, could be linked to cellular aggregation.

Cultivation
Cultures of bacteria-free Parachlorella kessleri (Chlorella vulgaris) strain where n is the rotations per minute (rpm) of the stir bar, D i is the length of the stir bar (m), ρ is the liquid density of the media (kg·m −3 ), and μ is the dynamic liquid viscosity at 22.3˚C (Pa·s). The Reynold's numbers were 0, 18,844, 26,696, 56,533, and 69,095. Based on these calculations, all turbulence-treated cultures reported here were within the turbulent, rather than laminar flow, range [27]. Experimental treatments were performed in duplicate and monitored for a period of ten days, after which, growth and flocculation efficiency were calculated. In addition, cultures were examined microscopically and tested for superoxide, hydrogen peroxide and polysaccharide content.

Cell Density and Flocculation Efficiency
Experimental cultures were stirred for 60 seconds, and then the OD 680 was measured immediately. Cultures were left undisturbed for a period of 20 minutes to allow for algal aggregates to settle. An aliquot of each sample was then obtained at a depth of 2.5 cm below the surface and the OD 680 was remeasured. Flocculation efficiency was calculated according to the following equation (Equ-Journal of Environmental Protection ation (2)) modified from Oh et al. [35]:

Hydrogen Peroxide Assay
Hydrogen peroxide concentrations were determined based on the method described by Li et al. [37] for the macroalgae Corallina officinallis. First, a 28 mL aliquot was transferred to a 50 mL conical tube and centrifuged at 2500 rpm for 8 minutes to produce a pellet weighing approximately 0.2 g (wet weight). Pellets were resuspended in 3 mL of a 0.1% solution of trichloroacetic acid (TCA) and placed in an ice bath for 3 minutes. Once chilled, a 1 mL sample was placed into a sterile 2 mL Corning ® polypropylene cryogenic vial with approximately 1.5 g of 0.5 mm zirconium/silica beads (BioSpec Products, Bartlesville, OK, USA) and bead-beaten for 10 seconds with three repetitions on high speed. Homogenate was placed on ice for another 3 minutes and then transferred to a 1.5 mL microcentrifuge tube. Samples were centrifuged at 8000 rpm for 8 minutes. A 0.5 mL aliquot of the resulting supernatant was combined with 0.5 mL of a potassium phosphate buffer at pH 7.0 and 1 mL of a 1 M solution of potassium iodide. After a 1 hr incubation in the dark, the absorbance was measured at 390 nm, compared against a standard curve, and expressed as μmol/g of wet weight.

Polysaccharide Assay
Polysaccharides, soluble extracellular and bound intracellular, were extracted and purified according to Yang et al.

Statistical Analysis
Statistical analyses were performed using SAS/STAT® statistical software. A two-way ANOVA with a randomized block design was used to test the effects of stress treatments at a significance level of p < 0.05. For each of the three stress treatments, two separate trials were completed at different times. Each trial was performed in duplicate. Since stress treatment trials were performed at different time points, a randomized block design was used to account for variations within the data associated with changes experienced between time points, in which stress treatments were considered fixed effects and trials were considered random effects. In addition, a Tukey test (p < 0.05) was performed to group homogenous means within each stress treatment, in which letters above bars in graphs represent the results of the Tukey analysis. All bars labelled "A" are statistically similar to each other as are all "B" or all "C". "A" vs. "B", "A" vs. "C", etc. are statistically different. Journal of Environmental Protection

Culture Growth
Algae were treated over a broad range (0 -1 mM) of glyoxylate concentrations in order to simulate nutrient deprivation. As seen in Figure 1(a), at lower concentrations glyoxylate had little effect on growth. A statistically meaningful effect on growth was found in a positive linear correlation (R 2 = 0.3755) with glyoxylate concentration and in an 8% increase (p = 0.0532) at the highest concentration (1 mM). As seen in Figure 1(b), increasing day length had an inhibitory effect on growth, in which growth decreased as day length increased from 8 to 24 hours (R 2 = 0.4034). The p-value (p = 0.1047) showed no statistical significance.
As seen in Figure 1(c), a significant relationship between turbulence treatment and growth was observed (p = 0.0039). Most notably, growth of cultures under a turbulence level of 300 rpm decreased by 15%. Higher turbulence showed little to no effect, which was confirmed by a Tukey test. Linear correlation between growth and increasing turbulence was lacking (R 2 = 0.0577).

Flocculation Efficiency
Cellular flocculation causes aggregations to form, which subsequently fall out of suspension due to their increase in weight [39]. Flocculation efficiency measures the proportion of the culture that has aggregated by calculating the difference in optical density pre-and post-settling of flocs. High flocculation efficiency values indicate a flocculation increase whereas zero indicates no effect. Both single cells and cellular aggregates were apparent within each of the cultures examined, as observed under the optical microscope ( Figure 2); flocculation efficiency was used as quantitative method for measuring relative aggregate frequency in response to stressors. The data followed a strong positive linear trend (R 2 = 0.5272).
A correlation between daylength and flocculation efficiency was not apparent (p = 0.3587, R 2 = 0.1805, Figure 3(b)); turbulence, on the other hand, induced a strong effect on flocculation efficiency which was both visually apparent and statistically significant (p ≤ 0.0001, Figure 3(c)). Within the lowest turbulence level  is expressed as a percent change from control ± standard deviation with n = 4. of 300 rpm the occurrence of flocculation tripled, while cultures subjected to turbulence levels of 900 rpm experienced an unexpected decrease in flocculation efficiency. The 425 and 1100-rpm treatments were statistically similar to each other and the control and 900-rpm treatment (as determined by Tukey's), and thus had much lower values of flocculation efficiency than the 300-rpm treatment.

Superoxide Production
The effect of glyoxylate, photoperiod changes, and turbulence stressors on superoxide and hydrogen peroxide concentrations were measured as indicators of cellular stress. After reduction of NBT by superoxide, a dark blue formazan precipitate was easily visualized within cells as small blue spots and quantitated at A 570 .
A small but significant increase in NBT concentration with glyoxylate treatment was observed (Figure 4  Slight increases in superoxide were also noted with the 425 and 1100-rpm treatments. The data did not show a linear correlation (R 2 = 0.000).

Discussion
The results shown here demonstrate that there are differences in cellular responses to different stressors and that they must be evaluated on a case-by-case basis. Glyoxylate treatment and nitrogen deprivation have each been shown to inhibit cell growth of microalgae [6] [21] [40]. In the study reported here, however, growth was not inhibited but slightly enhanced. Yang et al. [6], found that only the highest glyoxylate concentration of 1.25 mM arrested cell growth. The discrepancy between that study and this is unclear. It is possible that inhibition is masked by early enhancement of growth as a consequence of the increased photosynthetic activity that glyoxylate causes. Bergman [41] did show that glyoxylate stimulated growth in three cyanobacterial species. We found a significant increase in superoxide at the two highest glyoxylate concentrations. Menon et al. [42] found that Parachlorella kessleri with higher levels of ROS also had the highest growth rates. Since ROS are notoriously short-lived, measurement of ROS concentrations can only provide a "snap shot" of cell status [42]. The lack of effect that we found for glyoxylate on cellular growth could be attributed to cellular recovery over the ten days.
Whereas, EPS can act as a carbon overflow mechanism Yang et al. [6] noted a step-wise increase in EPS and aggregate occurrence and size with glyoxylate. We only observed an increase in aggregation at the highest glyoxylate treatment (1.0 mM) and extracellular polysaccharides remained largely unchanged. Instead, a decrease in intracellular polysaccharides was noted, and the same was noted for total polysaccharide (data not shown).
The increase in flocculation efficiency without substantial EPS production with glyoxylate was unexpected, since previous reports have shown a link [6].
However, despite the lack of significant extracellular polysaccharide production, significant cellular aggregation did occur. Thin layers of polymeric coatings have been shown to increase cell stickiness and aggregation in microalgae [6] [43].
Since our cultures were supplied solely with atmospheric CO 2 , the increased rate of glyoxylate-induced carbon metabolism could deplete media of available carbon, effectively negating the glyoxylate effect. As such, cellular perception of a nitrogen-deprived state would dissipate.
Studies have indicated that increases in day length produce a proportional increase in growth until reaching the saturation point of the photosynthetic machinery [26] [44]. This would be expected for a culture initially, or as a consequence of high growth, limited by light availability. We report that above 8 hr, light was not limiting since additional light did not appreciably increase growth.
A negative linear trend between growth and daylength was noted with growth maximized in the 8-hour day length. In the diatom Nitzschia and green alga Spaerocystis, growth rates declined with increasing day length [44]. Krzeminska et al. [45] found that 24-hour day lengths stimulated growth of Botryococcus braunii and Scenedesmus obliquus, but a 12-hour day length was optimal for three Neochloris species investigated. With varying light intensities, growth of Parachlorella kessleri were all significantly greater under a light regime of 16:8 (light:dark) than 12:12 or 8:16 [26]. Microalgae grown under short day lengths have been shown to adapt by increasing light harvesting pigments [44] which may, in part, explain the ability of C. kessleri to grow at an elevated rate when cultured under short days.
Our hydrogen peroxide results suggest that Parachlorella kessleri was experiencing light oversaturation at 20 and 24-hour day lengths. Interestingly, superoxide production did not follow the same trend. Hastings et al. [46] noted that SOD activity increases during the day in the dinoflagellate Lingulodinium. The decrease in growth and increase of hydrogen peroxide suggests that the cells exposed to longer days were under stress. β-carotene and α-tocopherols, which actively work to dissipate excess energy and directly quench ROS, accumulate in response to UV-B, long day lengths, and high light exposures [11]; these are likely factors in the muted responses we observed after 10 days. Flocculation efficiency was not appreciably impacted by variations in day length. In previous studies, we observed that cells grown at high cell densities with light limitation adhere to the sides of the growth chamber (data not shown). At the much lower cell densities of this report light availability did not appear to be limiting. Therefore, cell adherence due to light limitation was denecessitated.
Di Pippo et al. [9] found that higher irradiances of 60 and 120 μmol photons m −2 ·s −1 are required for biofilm formation including the 90 μmol photons m −2 ·s −1 used in this study. In addition, EPS has been shown to provide protective functions related to high irradiance, such as quenching ROS and protection against UV-B radiation [5] [8]. It was anticipated that an increase in day length would increase EPS production. An increase in extracellular polysaccharide was observed for a change to a shorter (8 and 12-hr) or longer (24-hr) day length, however, intracellular polysaccharide content did not change.
Wolfstein and Stal [47] found that polysaccharide secretion occurred as an overflow mechanism during low temperature stress, as cells reverted to a resting state. Alternative to the carbon overflow mechanism seen for excess light exposure, reversion to a resting state can cause carbon flux by shutting down the pathways through which carbon is typically funneled for metabolic processes.
Thus, carbon builds up within the intracellular spaces and must be diverted extracellularly as EPS. Reduced day lengths, perceived as a seasonal change [44], may also induce EPS production. If shortening the day length were perceived then a reduction in growth would be expected, and thus activation of EPS overflow. Our results confirm that large quantities of EPS are not required for aggregation, and subsequent biofilm formation, in C. kessleri.
Despite the frequency in which microalgae naturally experience turbulence, and the importance to environmental studies and commercial algae operations, research on this topic is fragmented [29]. Stirring produced a vortex in cultures subjected to 425, 900, and 1100-rpm, which was not seen in at 300-rpm. Regardless of vortex formation, Reynold's numbers were well above 2000, the laminar flow threshold, and thus produced a more irregular and disordered turbulent flow [27]. In the 300-rpm treatment, the consistency of stirring was compromised over time by a progressive increase in friction between the stir bar and the container bottom; the 300-rpm rate did not prevent settling of cellular debris.
Leupold [29] showed that growth of Parachlorella kessleri increases in response to shear stress, although Hosaka et al. [30] and Mitsuhashi et al. [30] noted that growth decreased initially under short-term treatment. Leupold et al. [29] also noted that growth rate depended on the magnitude of shear stress applied, increasing with rate but hitting a threshold then decreasing at higher rates.
Photosynthetic activity followed a similar trend but recovered after treatment ceased; in agreement with both the results noted by Hosaka et al. [48] and Mitsuhashi et al. [30]. The positive responses to turbulence are true for all species, but detriments are dependent upon the morphological characteristics of the alga [29] [49].
In the 300-rpm treatment cells were sinking to the bottom, and disrupting the motion of the stir bar. Several studies have noted that microalgae will naturally settle out of solution, increasing aggregation/autoflocculation, but not creating a firm attachment to the surface on which they settle [29] [50]. Any contact, but especially repeated encounters with the stir bar, would greatly increase the likelihood of mechanical damage. Growth was significantly decreased at 300 rpm. With greater turbulence, the positive and negative aspects of stirring appeared to have been balanced, as indicated by growth rates that remained relatively unchanged. At 900 rpm, the balance favored growth.
Flocculation efficiency was significantly increased at 300 rpm. At higher turbulence, flocculation efficiency decreased; in particular, the 900-rpm treatment had a negative % flocculation efficiency. It is worth noting that the treatment with the lowest growth rate had the highest measurement for flocculation efficiency (the 300-rpm treatment). In the 900 and 1100-rpm treatments a ring of cellular aggregation formed on the glassware directly below the water-air interface. This location may have served two purposes: one, this location was furthest from the stir bar, where the cell would experience a reduced turbulence and possibility of mechanical damage, and two, increased access to CO 2 .
Increased flow velocities will decrease the size of the cells' boundary layer, increasing cell: cell and cell: surface collisions [1]. Aggregation will occur until the shear becomes great enough to overcome cellular attachment [1]. Johnson and Wen [50] noted that microalgae will form a firm attachment, only when subjected to some turbulence.
Extracellular polysaccharide concentrations were elevated by turbulence.
Dayananda et al. [51] found that Botryococcus had greater EPS production with turbulence treatment, while Cordoba-Castro et al. [52] noted a linear relationship between turbulence rate and EPS secretion in Scenedesmus obliques. Although the mechanism has not been resolved, several studies have noted that EPS increases viscosity and thus decreases the impact of turbulence [21].
It would be expected that with similar levels of EPS, that aggregation in lower turbulence treatment would be less. Di Pippo et al. [1] noted that biofilm formations were disrupted by the detachment of flocs, or aggregates, by increasing circulation rate. This agrees with our results in that despite the higher collision frequency in treatments with higher turbulence flocculation efficiency was low.
The finding that extracellular polysaccharide concentrations remained consistent shows that increased polysaccharide secretion was not the only factor in the mechanism for attachment.
Rodríguez et al. [32] reported that oxidative stress increased during turbulence. Interestingly, ROS production depended greatly on the level of shear stress. Superoxide was significantly increased by the 300-rpm treatment, correlating with the decreased growth and increased flocculation efficiency. In contrast, at 900-rpm ROS was similar to that of the control. Our results suggest that cells undergoing mechanical stress, at 300-rpm do not acclimate.
Overall, the effect of the stressors investigated elicited strikingly varied res-ponses in EPS production and biofilm formation. The response of P. kessleri showed variation among the environmental stressors tested. These results were unexpected in that we set out to confirm what was reported in the literature then develop an effective harvesting method for algae grown for biofuels but we found a much more complex system. This body of research demonstrates the complexity of stress detection and response in these organisms.

Funding
This publication was made possible by funding from Appalachian State University and the USEPA grant (SU83471501). Its contents are solely the responsibility of the grantee and do not necessarily represent the official views of the USEPA. Further, USEPA does not endorse the purchase of any commercial products or services mentioned in the publication.