Introduction

Environmental conditions play a major role in determining the content and composition of lipids produced by microalgae [39]. Particularly important is the effect of growth conditions on the relative rates of cell division and the production of storage lipids (i.e. triacylglycerides) [20]. Driven to adapt over a wide range of growth conditions, many species have developed the ability to efficiently modify lipid metabolism in response to various external stimuli [14, 50, 53]. For example, the availability of nutrients such as phosphorus [26, 35, 44], nitrogen [5, 36, 43] and silicon [38, 62] has been shown to influence both lipid quantity and composition in many microalgae. Other environmental factors found to influence lipid composition include temperature [6], salinity [33, 49], light intensity [12, 33, 56], and light cycle [8, 48]. The ability to influence lipid metabolism in microalgae presents a commercial opportunity for the production of lipid-based products such as biofuel, oleochemicals and nutritive oils [58].

Triacylglyceride (TAG) oils are important lipids with application to the production of biodiesel [41] and nutraceutical oils [28]. The commercial viability of producing TAG-based products from microalgae is dependent on the productivity and cellular concentration that can be obtained. Nitrogen deprivation (N-deprivation) is a proven strategy to maximise the concentration of TAG in microalgae [20, 37, 50], and the ability of different algae to accumulate TAG in response to N-deprivation has been a key factor in the screening and selection of commercially suitable strains [5, 11, 21, 37]. When microalgae are N-deprived and unable to produce new proteins for growth, biosynthetic metabolism is directed towards producing and accumulating TAG as a stored form of carbon and energy [20, 50]. These TAGs are deposited as densely packed lipid bodies located within the cytoplasm of the microalgae [9].

The accumulation of TAGs during N-deprivation enables continued utilisation of light to produce new biomass from CO2, however, at a reduced overall productivity compared to growth under optimal conditions [42]. Strategies to balance this trade-off between biomass production and TAG accumulation are generally variations of a two-stage process whereby biomass concentration is increased prior to the depletion of nitrogen [5, 37]. Efforts to optimise N-deprivation growth regimes are ongoing, and have to date been primarily focussed on maximising TAG productivity. However, the effect of physical changes to the cells resulting from N-deprivation on the process efficiency of recovering of intracellular lipids from the biomass is still poorly understood.

The cell wall of microalgae and its response to changes in the growth environment is an important consideration in species selection considering it is the main barrier to the recovery of intracellular lipids [59]. The ability of microbial cells to resist mechanical rupture (e.g. by high pressure homogenisation) has been linked to the mechanical strength of the cell and the thickness of the cell wall [7, 29]. An increase in cell wall thickness of up to 70% has been reported for Chlorella emersonii when grown in a hypersaline culture, suggesting that cell wall thickness will vary with changes in growth conditions [30]. In another detailed study, N-deprivation was shown to have large effect on the internal cell structure of Nannochloropsis gaditana; however, changes to the cell wall or the mechanical strength of the cells were not investigated [43]. Similarly, the thickening of cell walls in Symbiodinium spp. from N-deprivation has been previously reported without further investigating its mechanical resistance towards disruption [23, 57]. While the effects of N-depletion on cell wall thickness have not been investigated in relation to cell disruption, a previous study by Van Donk et al. attributed the reduced digestibility of N-deprived phytoplankton by grazers to morphological changes in the cell wall [52], suggesting that an increase in the robustness of cell walls may occur during N-deprivation.

The aim of this study is to investigate the effect of N-deprivation on the cell size and cell wall thickness of three biotechnologically important marine microalgae, namely Nannochloropsis sp., Chlorella sp. and Chlorococcum sp., and to explore any correlation of these properties with the mechanical strength of the cells and their ability to resist mechanical rupture by high pressure homogenisation [15, 16, 19, 46]. Transmission electron microscopy (TEM) was used to investigate changes in cell wall thickness resulting from N-deprivation. Atomic force microscope (AFM) was used to probe the mechanical strength of individual cells. AFM enables force measurements of cells in the native aqueous environment to better understand the physical response of a cell to external stresses [4, 27, 54]. High pressure homogenisation was chosen as the method for mechanism cell disruption as it is a proven unit operation at scale and has been recently demonstrated to be energetically feasible for large-scale microalgal processing [60]. Here, we report for the first time the effect of N-depletion on the cell wall thickness and mechanical strength of microalgae in relation to their susceptibility to mechanical cell rupture.

Materials and methods

Microalgae cultures

Cultures of Chlorella sp., Chlorococcum sp. and Nannochloropsis sp. were maintained in flasks with a flat surface area of 25 cm2 (Corning Incorporated, Corning, NY) at 20 °C under a continuous low photon flux intensity of 6–10 μmol m−2 s−1 provided by white fluorescent lights. Cells were maintained in a modified ‘f-medium’ with nutrients and trace elements in synthetic seawater [13], with compositions as previously described [31]. Subcultures were prepared every 3–4 weeks.

To perform controlled experiments involving N-deprivation, 1.5 L of each species was first grown under N-replete conditions in aerated 2 L Schott bottles at 20 °C in a light:dark cycle of 12:12 h with a photon flux intensity of 60–70 μmol m−2 s−1. The aeration provided both a source of carbon and agitation for the cells in culture. Cells were harvested at the stationary phase via centrifugation at 5000g for 15 min at 20 °C (Beckman Coulter Avanti 30 bench-top centrifuge with a F0685 fixed angle rotor) and resuspended in 3–4 L of fresh medium without NO3 −. The resuspended cultures were then equally split into N-replete (NR) and N-deplete (ND) aerated cultures, whereby 5 mM NO3 − and 0.5 mM NO3 − were added, respectively. These cultures were then allowed to grow for 7–10 days before harvesting by centrifugation as above. The small amount of nitrate added to the ND media was included to reduce any possible shock effects, and was consumed by the cells within 2 days. The concentration of NO3 − in the NR cultures was not growth limiting within the time of the experiment, as confirmed by a previous study [28].

Image analysis by transmission electron microscopy (TEM)

Cells were pelleted in 1.5 ml Eppendorf tubes and dispersed in 2.5% glutaraldehyde in PBS for 2 h at room temperature after discarding the supernatant. The cells were then rinsed three times in fresh buffer for 10 min each before post-fixing in 1% osmium tetroxide in buffer for 1 h. The cells were again rinsed three times in fresh buffer for 10 min each, before being dehydrated in increasing concentrations of anhydrous ethanol of 10, 30, 50, 70, 90 and 100% v/v for 15 min each step. Following dehydration, the cells were infiltrated with increasing concentrations of LR White resin in ethanol consisting of 25, 50, 75 and 100% resin for 6 h each step. After a second change of 100% resin, the cells were embedded in fresh resin in gelatine capsules and allowed to gently sink to the bottom to form a loose pellet. The gelatine capsules were capped to exclude air and the resin polymerised in an oven at 60 °C for 24 h.

Embedded cells in blocks were sectioned with a diamond knife on a Leica Ultracut S microtome and ultra-thin sections (90 nm) were collected onto formvar-coated 100 mesh hexagonal copper grids. The sections on grids were sequentially stained with saturated uranyl acetate for 10 min and Triple Lead Stain for 5 min [40] and viewed in an FEI Tecnai Spirit transmission electron microscope at 120 kV. Images were captured with a Gatan Eagle camera at a resolution of 2048 × 2048 pixels. To minimise bias towards the hypothesis examined in this paper (i.e. N-depletion increases cell size and cell wall thickness), samples were fixed and images were captured without the knowledge of the growth conditions of the cell samples.

Analyses of cell sizes and cell wall thickness were performed using ImageJ, a public-domain image processing and analysis software [34]. The cell wall thickness was measured in 3 separate places on a single cell and a total of 30 individual cells per species were measured. Data collected for each species were subjected to unpaired two-tailed t tests for the difference in means (N-replete vs N-deplete) assuming unequal variance where a p value of less than 0.05 was considered to be significant.

Cell disruption by high pressure homogenisation

Cell disruption experiments were performed using a GEA Panda2K NS1001L bench top high pressure homogeniser (GEA Niro Soavi, Parma, Italy) with a nominal flow rate of 10 L h−1 and equipped with an RE+ valve. Microalgal suspensions were processed in a single pass through the homogeniser at pressures ranging from 30 to 150 MPa. Analyses of the recovered homogenates were performed within 3 h of homogenisation. All cell disruption tests were performed in biological triplicates.

Cell rupture was quantified by cell counting due to its accuracy and reproducibility [46]. The number of intact cells remaining after homogenisation was counted using a Neubauer improved hemocytometer (Laboroptik Ltd., Lancing, UK) with a 100 μm chamber depth. Cell counting was performed using an Olympus BX51 light microscope with a DP72 digital camera attachment (Olympus, Mt. Waverly, VIC, Australia). Cell counts were normalised between the control sample (unhomogenised cell count) and zero.

Force measurements by atomic force microscopy (AFM)

The mechanical stiffness of individual Chlorococcum sp. cells was determined using a colloidal probe in a series of cell indentations using an MFP-3D AFM (Asylum Research, Santa Barbara, CA, USA). Spherical silica beads of 50–80 μm in diameter (Thermo Fisher Scientific Inc., Waltham, USA) were glued to the apex of a rectangular silicon cantilever (Tap150-G, Budget Sensors, Sofia, Bulgaria). The spring constant of the cantilevers was determined to range from 10 to 15 N m−1 using the thermal method [22]. All measurements were performed at room temperature in a modified ‘f-medium’ in synthetic seawater with the presence or absence of NO3 − depending on the native environment of the measured sample [13].

The AFM fluid cell was fitted with a circular glass disk coated with a layer of 0.1% w/v poly-l-lysine (MW: 700–100 kDa, Sigma Aldrich, Australia) for cell immobilisation, onto which the cells were transferred to and allowed to settle in a single plane. Silica bead attachment onto the cantilever and visualisation of cells were performed using an inverted light microscope (Nikon Eclipse TE2000-U) attached to the AFM. The colloidal probe consisting of a silica bead attached to the cantilever was positioned above an immobilised cell that was considerably smaller than the diameter of the bead. A schematic of the experimental setup is shown in Fig. 1. The force applied to the cell was kept below 1 μN, which corresponded to an average indentation of less than 1 μm into the cells. This amount of indentation into the cell was less than 10% of the total height of the cells, a required condition to apply the Hertz model for data analysis.

Fig. 1
figure 1

A schematic of the experimental setup for the compression of a microalgae cell using a colloidal probe in an AFM (not to scale)

Full size image

The AFM records the cantilever detector photodiode voltage as a function of the piezo movement, ∆l using a linear variable differential transformer (LVDT). The raw photodiode voltage is converted to cantilever deflection, ∆d by scaling the photodiode detector sensitivity, which was determined from the slope of the force curve obtained by pressing the cantilever against the rigid substrate surface. The force, F, was then calculated by multiplying the cantilever deflection with the spring constant, k (Eq. 1).

$$F = k\Delta d$$
(1)

To obtain a force versus indentation curve, indentation (or deformation into the cell) was calculated by subtracting the cantilever deflection from the piezo distance (Eq. 2). The contact distance was assumed to be the point at which the deflection rose above the baseline deflection force level for each force curve [4].

$$\Delta z = \Delta l \, - \Delta d$$
(2)

A Hertzian contact model approximation was applied to the force versus indentation curve to obtain the Young’s modulus of individual cells. Cell stiffness is one of the most common parameters measured to investigate change in the mechanical properties of biological cells. AFM was used to indent Chlorococcum sp. cells grown in N-replete and N-deprived conditions to estimate the Young’s modulus in order to quantify the difference in their stiffness. The Hertz model [18, 24] describes the simple case of elastic deformation of two homogenous bodies touching under load. An inherent assumption for applying the Hertz theory is that the material being probed must be isotropic on the scale of the indentation, which is often not the case for living cells like microalgae. However, given that the objective of this work is to determine relative differences rather than absolute elasticity values, the Hertzian approximation is adequate. A sphere–sphere configuration for the indentation geometry is assumed in this analysis considering Chlorococcum sp. was found to have a circularity of >0.99 and the silica bead attached to the tip is spherical. The cells of Chlorella sp. and Nannochloropsis sp. are ellipsoid, meaning that the sphere–sphere model could not be applied. This issue could not be resolved as the orientation of the cells’ axes could not be controlled during the measurements. Thus, for Chlorococcum sp. cells, the relationship between the applied load, contact radii and the displacement of elastic bodies is given by Eq. (3):

$$F = \frac{4}{3}E^{*} R^{*1/2} \delta^{3/2}$$
(3)

where

$$\frac{1}{{E^{*} }} = \frac{{1 - v_{1}^{2} }}{{E_{1} }} + \frac{{1 - v_{2}^{2} }}{{E_{2} }},\quad \frac{1}{{R^{*} }} = \frac{1}{{R_{1} }} + \frac{1}{{R_{2} }}$$
(4)

and F (N) is the force applied to the cell, E (Pa) is the Young’s modulus in Pa, R (m) is the radius of the sphere, δ (m) is the indentation depth and v is the Poisson’s ratio. Subscripts 1 and 2 denote the silica bead (i.e. colloidal probe) and cell, respectively. The Young’s modulus, E 1 and Poisson’s ratio, v 1 for the silica bead were taken as 7.3 × 1010 Pa and 0.17, respectively [10]. Another limitation in using the Hertz model in measurements involving biological cells is that the Poisson’s ratio is not available. To determine the Young’s modulus for a cell, the Poisson ratio for Chlorococcum sp. was assumed to be 0.5 [47, 55]. For the purpose of comparing relative values of elastic modulus in this study, it is sufficient to assume incompressibility (i.e. v = 0.5) when fitting the model to experimental data. Data collected for both the N-replete and N-deplete samples of Chlorococcum sp. were subjected to unpaired two-tailed t tests for the difference in means assuming unequal variance where a p value of less than 0.05 was considered to be significant.

Results

To investigate the effect of N-deprivation on the cell size and cell wall thickness of the three microalgae, TEM images were taken (Fig. 2) and analysed. For all three microalgae, there was a statistically significant (p < 0.0001) increase in the mean cell size and cell wall thickness as a result of N-deprivation (Table 1). For Nannochloropsis sp. and Chlorococcum sp., both the cell size and cell wall thickness were approximately 25% greater in the ND cultures. Correspondingly, the ratios of cell wall thickness to cell size were also found to remain relatively constant for both NR and ND samples of Nannochloropsis sp. and Chlorococcum sp. For the ND Chlorella sp., the average cell wall was approximately 70% thicker and the average cell size only 20% larger. While there were highly statistically significant differences between the means of these parameters between NR and ND cultures, there was also considerable overlap between the populations (Fig. 3) which is expected given the cell-to-cell variability within a population, particularly between cells at different stages in the growth cycle.

Fig. 2
figure 2

Representative TEM micrographs of Nannochloropsis sp., Chlorella sp. and Chlorococcum sp. grown in N-replete and N-deplete conditions. Scale bars for the cell wall and whole cell micrographs for Nannochoropsis sp and Chlorella sp. indicate 100 nm and 1 μm, respectively. Scale bars for the cell wall and whole cell micrographs of Chlorococcum sp. represent 200 nm and 2 μm, respectively

Full size image
Table 1 Mean cell wall thickness and cell size of Nannochloropsis sp., Chlorella sp. and Chlorococcum sp. grown in N-replete (NR) and N-deplete (ND) conditions
Full size table
Fig. 3
figure 3

Histograms comparing the cell wall thickness and cell size of N-replete (NR) and N-deplete (ND) cultures of Nannochloropsis sp. (a, b), Chlorella sp. (c, d) and Chlorococcum sp. (e, f). Mean values are presented in the Table 1

Full size image

The elastic response of a cell to an external force is an indicator of its mechanical strength. To determine if there was any correlation between the observed increase in cell wall thickness resulting from N-deprivation and the actual mechanical strength of the cells, compression tests via cell indentation were performed on individual cells of Chlorococcum sp. using an AFM. A representative force versus indentation curve and subsequent Hertzian approximation of the Young’s modulus are shown in Fig. 4. As mentioned in “Materials and methods”, the point of contact is attributed to the rise in force above the baseline. In the inset plot (Fig. 4), one can see that force or load follows a linear relationship with the indentation raised the 3/2 power. This is consistent with describing these data with a Hertz model for this geometry. The retract portion of the AFM force curves sometime exhibits adhesion between the cell and the particle and, thus, the advancing portion of the AFM force curve was used for the Hertz analysis. The distributions of the Young’s moduli between the ND and NR populations (Fig. 5) overlap due to the broadness of the destructions attributed to both physical and biological variation between cells. However, the moments of the distributions reflect that there are large statistically significant differences between the means of the two populations (p = 0.0048). The mean Young’s modulus of the ND population (n = 44) was approximately 30% greater than the NR population (n = 42) (775 and 619 kPa, respectively).

Fig. 4
figure 4

A representative force versus indentation curve for a Chlorococcum sp. cell and a Hertzian sphere–sphere contact model approximation applied to the data (inset plot). Young’s modulus of a cell, E 2, is obtained through the slope of the curve that is represented by the term \(\frac{4}{3}E^{*} R^{*1/2}\) (see: Eq. 3). An R 2 of 0.996 was obtained for the curve represented here

Full size image
Fig. 5
figure 5

Histogram comparing the Young’s moduli of N-replete (NR) and N-deplete (ND) Chlorococcum sp. cells. Mean Young’s modulus for NR and ND cells is 619 ± 292 and 775 ± 200 kPa, respectively (difference between means p value = 0.0048)

Full size image

To test whether the observed changes in the average cell size, wall thickness, and for Chloroccoum sp. the cell mechanical strength had any effect on the susceptibility of the cells to mechanical rupture, populations of cells from NR and ND cultures were subjected to high pressure homogenisation at a range of pressures. Even though a statistically significant (p < 0.0001) increase in cell wall thickness was observed across all three species, no difference in cell rupture behaviour was observable in the data obtained (Fig. 6). Data for Chlorococcum sp. were not shown as the cells were very weak and were completely broken by 35–40 MPa for both NR and ND samples. The results show Nannochloropsis sp. to be the most resistant to mechanical rupture, followed by Chlorella sp. and, subsequently, Chlorococcum sp., consistent with findings reported in a previous study [46].

Fig. 6
figure 6

Cell rupture as a function of pressure for a single homogenising pass for a Chlorella sp. and b Nannochloropsis sp. grown in N-replete (NR) and N-deplete (ND) conditions. Error bars represent the standard deviation of triplicate experiments performed on separate algal cultures

Full size image

Discussion

The cell walls of the microalgae investigated in this study are similar to that of higher plants in that they consist of continuous structures of biopolymeric microfibrils [2]. In contrast, the coverings of some other microalgae are rather different, for instance the scaly/plate-like coverings of microalgae such as Isochrysis galbana and Tetraselmis suecica and the silica cell coverings of diatoms. For microalgae with plant-like cells walls, such as the ones examined here, the cell wall can be considered to be an elastic structure that functions to maintain cell integrity through a combination of mechanical strength and flexibility. The fundamental components of the cell wall in the species of microalgae studied here include a microfibrillar network within a gel-like matrix of proteins, hemicelluloses and pectins [1]. The microfibrils consist of chains of monosaccharide residues and cellulose. Cell walls of representative species from the genus Chlorella typically consist of 44–77% neutral sugars, 4–24% uronic acids, 2–11% proteins and 2–5% glucosamine [3]. The chemical composition of cell wall could further inform about the efficacy of pre-treatments to weaken the cell wall prior to mechanical cell disruption [16, 32, 45], but is beyond the scope of this present study.

The elastic moduli of Chlorococcum sp. determined here are comparable to other microorganisms with similarly structured cell walls. For example, a previous study using the AFM determined cells of the green microalga Scenedesmus dimorphus to have an elastic moduli of ca. 2 MPa in an aqueous environment [54]. Another eukaryotic microorganism, S. cerevisiae, also investigated using an AFM, was found to have an elastic modulus of 0.6 MPa, although some regions of the cell wall (i.e. the bud scar) had a significantly higher elastic modulus of 6 MPa [51]. In contrast, a much higher Young’s modulus of up to 22.4 GPa has been reported for diatoms that possess rigid silica cell walls [17]. None of these studies investigated the relationship between the Young’s modulus and cell wall thickness. The results of the current study show a positive correlation between the Young’s modulus of the entire cell and the cell wall thickness of Chlorococcum sp. In addition to increased cell wall thickness, it is possible that changes to the internal structure of the cells [43] that are not accounted here could have contributed to the increase in Young’s modulus. Regardless, for all three microalgae tested here, the increase in cell wall thickness and corresponding increase in the Young’s modulus of the cells were not sufficient to result in an observable difference in their ability to resist mechanical rupture through high pressure homogenisation.

The independence of the susceptibility of cells to mechanical rupture on N-deprivation is an important finding. It suggests that lipid accumulation can be induced by N-deprivation without resulting in extra difficulties in breaking the cell wall to recover the intracellular lipids. These results also demonstrate that it can be difficult to establish direct links between microstructural changes and observable changes in bulk processing behaviour as the relationship is not straightforward. An example of these difficulties can be found in the previous work of Lee et al. where an extrapolation of the mechanical strength of a single cell was used to estimate the specific energy required to disrupt a kg of biomass (i.e. by multiplying an assumed number of cells per kg of biomass) [25]. The estimated values derived from the microstructural information in the Lee et al. study turned out to be vastly higher than what is actually required [61].

Despite being drawn from a limited set of data, it is perhaps notable that the relative resistance to rupture of the different species was in reverse order to the average size of the cells. This was despite a trend of decreasing cell wall thickness with decreasing cell size, and in fact also a decreasing ratio in cell wall thickness to cell size (Table 1). Although the strength and elasticity of the cell walls of different algae are not equal, it nonetheless suggests that cell size is an important factor in determining the ability of cells to resist rupture when passed through the narrow valve gap in a high pressure homogeniser. This could also be a factor in the lack of increased resistance to rupture shown by ND populations, which in addition to having thicker cell walls were on average larger than the NR cells. The coincidental increase in cell size may have increased the susceptibility of the ND cells to rupture, counteracting the strengthening provided by the thicker cell walls. It is, however, important to also note that apart from cell size and cell wall thickness, other factors such as the overall composition of the cell wall may also contribute towards the susceptibility of the cell towards mechanical rupture. Combining these observations with other insights obtained from previous work could provide further insight into the dominant mechanism of cell breakage in a high pressure homogeniser. For instance, it has been observed that the percentage of cells ruptured does not change significantly as a function of passes through the homogeniser (ruling out biological variability within the population as a dominant factor) [46] and that cell rupture is independent of solids concentration [60].

Conclusion

Nitrogen deprivation was found to significantly increase the average cell size and wall thickness of Nannochloropsis sp., Chlorella sp. and Chlorococcum sp. The increased cell wall thickness was correlated to an increase in the mechanical strength of N-deprived Chlorococcum sp. However, the differences in microstructural properties between N-replete and N-deplete microalgae did not translate to a higher resistance to cell breakage when processed through high pressure homogenisation. This result is particularly important for the development of growth regimes optimised to maximise lipid productivity through nitrogen starvation, as it indicates that cell rupture is not likely to be significantly affected for these algae.