Comparison of Tetraselmis suecica Cell Disruption Techniques: Kinetic Study and Extraction of Hydrosoluble Compounds

Abstract

The optimization of cell disruption is a critical step in microalgal biorefineries. We used the same batch of Tetraselmis suecica culture to compare two mechanical cell disruption techniques, focusing on the extraction yield of water-soluble molecules. The conditions for high-pressure homogenization (HPH) studied were two passes at a moderate pressure of 300 bars. For ultrasound (US) treatment, we used an amplitude of 20% (equivalent to 100 W) for 25 min. These conditions were chosen on the basis of a preliminary screen of extraction conditions. HPH extracted proteins and pigments more efficiently than US, whereas US was superior for uronic acid extraction. Interestingly, the two methods had similar extraction yields for carbohydrates under the studied conditions. We also analyzed the kinetics of molecule release by considering the centrifugation time lag for HPH and applying a first-order kinetic model for US. HPH outperformed US in terms of the immediate extraction and release of molecules.

1. Introduction

Microalgae are a prospective source of valuable nutrients, including pigments, proteins, carbohydrates, and lipid molecules [1]. They use carbon dioxide (CO₂) at rates 10–50 times higher than those of terrestrial plants, produce biomass faster than land plants, and have the capacity to remove nitrogen, phosphorus, and heavy metals from water [2]. Despite ongoing research efforts to overcome issues potentially hindering the release of microalgal molecules, efficient downstream processing remains a major challenge for commercial-scale implementation [3]. Cell wall disruption is a crucial step in the efficient recovery of intracellular molecules and the fractionation of cell wall components [4]. Several cell disruption methods have been applied to microalgae, with various degrees of success. In cell disruption, it is essential to differentiate between “permeabilized cells” and “lysed cells.” The walls and membranes of the cells are partially ruptured in permeabilized cells, whereas the lysed cells are completely disintegrated [5]. Cell disruption methods can be classified as mechanical (bead milling, high-pressure homogenization, ultrasound, etc.) or non-mechanical (chemical and enzymatic) methods on the basis of the mechanisms used for cell disintegration [6].

The efficiency of cell disruption depends on the cell wall structure. Some microalgae can be easily disrupted by mild or more energy-efficient disruption techniques, facilitating the direct access of the solvent to intracellular compounds and thereby increasing the extraction efficiency [7]. The use of such methods ensures high product recovery, low operating costs, and a high degree of purity of the extract [8]. High levels of polysaccharides in the cell wall and the presence of a non-hydrolysable biopolymer composed of long ω-hydroxy fatty-acid chains render the cell wall rigid [9].

Tetraselmis suecica is a fast-growing marine green microalga that accumulates proteins, carbohydrates, lipids, and other molecules [10]. It can withstand a diverse range of salt concentrations [11] and is widely used in aquaculture as a feed for bivalves, shrimp larvae, rotifers, and brine shrimp [12]. Tetraselmis suecica is an ovoid unicellular flagellated green alga that is about 9–13 µm long and 7–8 µm wide. Its composition varies with the growing conditions, which influence the accumulation of the target components [13]. Tetraselmis suecica fractionation has been widely studied because this alga has a high protein content. However, these studies have highlighted difficulties in optimizing its valorization, particularly in terms of the selection of an efficient extraction technology.

High-pressure homogenization (HPH) is a mechanical cell disruption technique in which very high pressures (from about 200 to 1200 bars) are used to generate turbulence, liquid shear stress, and friction, leading to the release of the contents of the microalga. The efficiency of HPH depends on several parameters, including the operating pressure and the number of homogenization cycles [14,15]. This continuous system based on the use of algal suspensions or slurries is easy to scale up. During this process, the biomass is forced through a narrow nozzle outlet by a high-pressure pump [16]. The cell suspension passes radially through a valve, strikes a ring, and then passes through another valve or is discharged. The mode of action of HPH remains unclear, but it is thought that the combination of a high shear force, cavitation, and a sudden decrease in pressure between the nozzle and the external environment leads to cell disruption [8]. Magpusao et al. [17] performed studies to optimize pressure levels for the HPH treatments of Tetraselmis sp. They found that a pressure of 300 bars caused substantial disintegration, with complete disintegration occurring at 600 bars. However, few studies have reported the effect of HPH on the disruption of Tetraselmis suecica. Spiden et al. [18] reported that lower pressures were sufficient to cause high levels of Tetraselmis suecica cell disruption by HPH, with a degradation of metabolites occurring at pressures above 90 bars, according to measurements of UV absorbance. Safi et al. [13] showed that the rate of Tetraselmis suecica cell rupture increased with the homogenization pressure: cells were resistant at pressures up to 400 bars, and the effect of HPH began to occur at 600 bars.

In addition, previous research on microalgae has demonstrated enhanced cell disruption with varying numbers of passes. For instance, Zhang et al. [19] found that 10 passes resulted in the highest extraction yield of proteins and carbohydrates from Parachlorella kessleri, showcasing the importance of exploring different pass numbers. Delran et al. [20] studied the effect of the pressure applied and the number of passes on the destruction of Tetraselmis suecica cells and reported that a single pass at a pressure of 400 bars was sufficient for protein extraction. None of these studies investigated the effect of a pressure of 300 bars on Tetraselmis suecica cell disruption, and it is therefore possible that the threshold for disruption lies closer to 300 bars than to 400 bars. At such a low pressure of 300 bars, the effect of the number of passes was not investigated. Importantly, none of these studies performed diffusion kinetic modeling on the extracted molecules.

Unlike HPH, ultrasonication (US) is a physical process in which sound waves with frequencies above 20 kHz are used to induce acoustic cavitation in a liquid medium [5]. These waves generate a liquid shear that moves from the metallic tip to the concentrated cellular suspension [21], causing the formation of microbubbles, which increase in size and then collapse, creating cavitation conditions. Violent bubble collapse increases the temperature to above 4000 °C and the pressure by hundreds of atmospheres [6]. Various parameters, including temperature, ultrasound frequency, cell concentration, acoustic intensity, number of cycles, wave amplitude, viscosity, mode used, and treatment time, should be controlled during this process [4,22]. Most studies have focused on lipid extraction from Tetraselmis suecica by US [23,24,25], but a few have considered the effects of US parameters on cell rupture and the release of water-soluble molecules. Delran et al. [26] studied the effects of US frequency, power, and treatment time on the destruction of Tetraselmis suecica cells and reported that low-frequency US (20 kHz) treatment allowed the extraction of 90% of proteins and 70% of carbohydrates.

The kinetics of Tetraselmis suecica protein release by US were investigated by Delran et al. [26], who studied the effect of the treatment time on the protein extraction yield without using a kinetic model. Suarez Garcia et al. [27] investigated the kinetics of protein release from Tetraselmis suecica by bead milling and used a first-order kinetic model. Halim et al. [23] proposed a kinetic model for Tetraselmis suecica cell lysis by HPH and US, in which the concentration of lysed cells was determined by hemocytometry. No kinetic model has yet been established for the extraction of water-soluble proteins from Tetraselmis suecica by US, and there have been no kinetic modeling studies of biomolecule release from Tetraselmis suecica during US treatment.

Major studies of the disruption of Tetraselmis suecica cells have been performed independently for different techniques. In this study, we compared HPH and US on the same batch of cells. It is not appropriate to compare cell disruption between different batches of Tetraselmis suecica because the morphology and physiological state of the biomass received may differ between batches. Mear et al. [28] identified various differed morphological properties and physiological states in Tetraselmis chui cultures obtained from the same supplier and grown in the same medium, highlighting disparities in physiological states between different batches. Cell wall rigidity and the biochemical composition of the biomass in the two batches they studied were not comparable, and the observed differences had direct impacts on cell disruption and protein recovery.

The primary objective of this study was to determine the most effective cell disruption technique and conditions for extracting hydrosoluble compounds from Tetraselmis suecica after centrifugation. Specifically, our aim was to compare the performance of two mechanical techniques, HPH and US, across various conditions. This comparison will be based on two critical factors:

• Yield of extracted molecules—Quantifying the extracted molecules in the supernatant after cell disruption.
• Kinetic release rate—Determining the time required for the extraction process, reflecting the efficiency of extracting a defined biomass quantity.

The selection of the optimal cell disruption technique based on these factors was performed using two different batches of Tetraselmis suecica. We then compared the most promising conditions of each technique within the same batch of cells. This study will facilitate the comparison of techniques and the optimization of processes, thereby increasing efficiency across diverse scientific and industrial applications.

2. Materials and Methods

2.1. Microalgae

Tetraselmis suecica was purchased from Inalve (Nice, France). It was produced as a biofilm (patented rotating algal growth system, WO2021180713A1) and had a dry matter content ranging from 15% to 20%, depending on the batch. Fresh biomass was received two days after harvesting as a concentrated paste. It was stored at 4 °C in the dark, and diluted to a concentration of 5% dry weight (DW). A portion of the biomass was freeze-dried for further characterization of its total biomolecule content to prevent rapid degradation. The rest was diluted 1:3 for fractionation. Cell disruption and centrifugation were performed within 48 h of biomass reception.

2.2. Reagents

The Bradford assay kit (including BSA standards) was purchased from Thermo Fisher Scientific, Illkirch-Graffenstaden, France. All the chemicals used were of analytical grade and were purchased from Sigma Aldrich, Saint-Quentin-Fallavier, France.

2.3. Treatments

Before cell disruption experiments, the fresh paste was diluted 1:3 with distilled water by gentle shaking at 150 rpm for 30 min on an orbital shaker. After each disruption, the supernatant was recovered by centrifugation at 10,000× g for 10 min at 20 °C in a 6–16 K centrifuge (Sigma Laboratory Centrifuges, Osterode am Harz, Germany) and subjected to biochemical analysis.

2.3.1. Control

The biomass dilution, corresponding to a non-mechanical water extraction, was considered a blank for comparison with the other treatments in this study [20,29].

2.3.2. Ultrasonication (US)

Ultrasonic extraction was performed with a Sonics Vibra-Cell ultrasonic processor probe–500 W (Thermo Fisher Scientific, Illkirch-Graffenstaden, France), operating at a frequency of 22 kHz. Sonication was performed on 250 mL of biomass at two amplitudes, 20% (100 W) and 70% (350 W), in a plastic beaker, with a 13 mm-diameter probe placed 5 cm below the surface. The chosen tip depth in the volume was based on the recommendation of the manufacturer. The treatment was performed in cool conditions (10 °C) and in pulsed mode, with cycles of 5 s of US and 20 s of resting time to prevent overheating of the sample. The suspension was irradiated for various times (5, 10, 15, and 25 min), and several samples were collected at different time points.

2.3.3. High-Pressure Homogenization (HPH)

This treatment was performed with an APV-1000 Lab homogenizer (Søborg, Denmark), with a pressure range of 0 to 1000 bars and a flow rate of 22 L/h. All experiments were performed at room temperature (approximately 20 °C) without a cooling system. The temperature at the HPH outlet never exceeded 40 °C in any of the conditions used. A volume of 250 mL was used for experiments relating exclusively to this technique, whereas a volume of 300 mL was used for the experiments comparing HPH with US. The impact of HPH was assessed under three different sets of conditions, as outlined in

Table 1. HPH operating conditions.

Condition	Pressure in Bars	Number of Passes
1	300	2
2	300	5
3	600	2

.

2.4. Extraction Kinetics

A kinetic study was also conducted to investigate the release of metabolites during US and HPH.

2.4.1. Ultrasonication (US)

A kinetic model was used to describe the diffusion of proteins and carbohydrates during Tetraselmis suecica cell lysis by US.

In this model, first-order release kinetics for proteins or carbohydrates were used to calculate the kinetic constant k as follows:

$${-\mathrm{r}}_{\mathrm{molecules}}=-\frac{{\mathrm{dC}}_{\mathrm{molecules}}}{\mathrm{dt}}=\mathrm{k}.{\mathrm{C}}_{\mathrm{molecules}}$$

(1)

$${\mathrm{C}}_{\mathrm{molecules}}={\mathrm{C}}_{0,\mathrm{molecules}}-{\mathrm{C}}_{\mathrm{released} \mathrm{molecules}}$$

(2)

where

• R_molecules—Diffusion rate in mg/mL.s;
• C—Concentration of the molecule remaining in the cell in mg/mL;
• C_{0, molecules}—Total concentration of the molecule in the initial biomass in mg/mL;
• C_{released molecules}—Concentration of the molecule released into the supernatant in mg/mL;
• k—Molecule extraction rate constant in s⁻¹;
• t—Treatment time for US in minutes.

The primary objective of kinetic modeling is to establish a quantitative relationship between the concentration of molecules released into the supernatant and the duration of US (treatment time). This relationship is derived from the solution of Equation (1) as follows:

$${\int }_{{\mathrm{C}}_{0,\mathrm{molecules}}}^{{\mathrm{C}}_{\mathrm{molecules}}}\frac{{\mathrm{dC}}_{\mathrm{molecules}}}{{\mathrm{C}}_{\mathrm{molecules}}}=-\mathrm{k}{\int }_0^{\mathrm{t}}\mathrm{dt}$$

(3)

to give

$$-\ln \frac{{\mathrm{C}}_{\mathrm{molecules}}}{{\mathrm{C}}_{0,\mathrm{molecules}}}=\mathrm{kt}$$

(4)

Replacing Equation (2) in Equation (4) yields:

$$-\ln \frac{\left({\mathrm{C}}_{0,\mathrm{molecules}}-{\mathrm{C}}_{\mathrm{released} \mathrm{molecules}}\right)}{{\mathrm{C}}_{0,\mathrm{molecules}}}=\mathrm{kt}$$

(5)

We define two dimension-less variables to simplify the above equations.

The fraction of molecules released, X, is calculated as follows:

$$\mathrm{X}=\frac{{\mathrm{C}}_{\mathrm{released} \mathrm{molecules}}}{{\mathrm{C}}_{0,\mathrm{molecules}}}$$

(6)

The time fraction, ζ, is calculated as follows:

$$\mathsf{\zeta }=\frac{\mathrm{t}}{{\mathrm{t}}_{\max }}$$

(7)

where t_max is the maximum treatment duration, which was 25 min of US.

Equation (5) can then be simplified as follows:

$$-\ln (1-\mathrm{X})=\mathrm{kt}$$

(8)

2.4.2. High-Pressure Homogenization (HPH)

The biomolecule extraction kinetics of Tetraselmis suecica were also determined on the basis of the centrifugation time lag, defined as the time between the completion of HPH and the beginning of centrifugation. This study was performed to ascertain whether the released molecules would continue to diffuse out of the residual solid matrix after cell rupture. We studied six extraction times after HPH extraction: 5 min, 20 min, 35 min, 1 h, 2 h, and 3 h. The extraction kinetics were studied for the three conditions shown in Table 1. This is the first study, to the best of our knowledge, to have considered the effect of the time lag to centrifugation. The kinetics of biomolecule diffusion into the aqueous supernatant were monitored for proteins, neutral carbohydrates, and pigments, including chlorophyll a, chlorophyll b, fucoxanthin, lutein, and β-carotene.

2.5. Biochemical Analysis

A biochemical analysis was performed to evaluate the biomass at different stages of treatment. We analyzed fresh biomass (dry weight and ash content), freeze-dried initial biomass (total proteins, total carbohydrates, and uronic acids exclusively for comparisons of HPH and US), diluted biomass (dry weight), freeze-dried disrupted biomass (total pigments exclusively for comparisons of HPH and US), and the supernatant (dry weight, ash content, total proteins, soluble proteins, carbohydrates, uronic acids, and pigments).

Dry weight was determined gravimetrically by heating the sample in an oven at 103 °C until its weight stabilized. The ash content was determined by calcination at 550 °C for 12 h.

For the total protein biomass analysis and extraction yield calculation, proteins were analyzed by the Kjeldahl method with a Kjeltec 8400 automatic analyzer (Nanterre, France) using a standard conversion factor (nitrogen to protein) of 5.95 [30]. For the analysis of soluble proteins and the comparison of disruption methods, the Bradford assay was used, with bovine serum albumin (BSA) as the standard [31].

The carbohydrate content was determined according to the protocol of the National Renewable Energy Laboratory (NREL) [32]. In brief, 25 mg of freeze-dried biomass was hydrolyzed with 250 µL of 72% sulfuric acid for 1 h at 30 °C in a water bath. We then added 7 mL of ultrapure water, and the biomass was hydrolyzed for 1 h in an autoclave at 121 °C. After cooling, the acidic hydrolysate was neutralized with calcium carbonate, filtered through a 0.22-micron filter, and analyzed. The monosaccharide distribution was determined with an HPIC DIONEX ICS 3000 DC-EG (Illkirch-Graffenstaden, France) with a Carbo-Pac PA-1 column, a post-column with 300 mM NaOH as the elution buffer, and an AS3000 automatic sampler. The device was equipped with a pulsed electrochemical detector.

Uronic acids were analyzed by the Blumenkrantz titration method [33]. The biomass was hydrolyzed for 5 min at 100 °C in a solution of sodium tetraborate (a reactant) in sulfuric acid. Once cooled, a basic metahydroxybiphenyl solution was added, and the absorption was recorded at 520 nm with a BMG-Labtech Spectrostar-Nano spectrophotometer (BMG LABTECH SARL, Champigny s/Marne, France). The concentration was determined by comparison with a calibration curve obtained with galacturonic acid as the standard. The uronic acid content was reported on a dry weight basis as follows:

$$\%\mathrm{Uronic} \mathrm{acid}=\frac{\mathrm{Concentration}\times {\mathrm{V}}_{\mathrm{reactant}+{\mathrm{H}}_2{\mathrm{SO}}_4}}{{\mathrm{m}}_{\mathrm{sample}}\times \%\mathrm{DW}}\times 100$$

(9)

Total pigments (fucoxanthin, lutein, chlorophyll a, chlorophyll b, and β-carotene) were extracted from the freeze-dried biomass according to a slightly modified version of the NREL protocol [34]. The fresh biomass was disrupted by HPH, with two passes at a pressure of 300 bars, and immediately freeze-dried. Purified water (100 µL) was added to 15 mg of the dried lysate, which was then vortexed for 15 s and stored on ice for 30 min to limit biological activity during rehydration. We then added 0.5 mL of methanol and vortexed the mixture for 15 s before adding 0.5 mL of chloroform and vortexing for another 15 s. The samples were centrifuged at 5000 RCF, and the supernatant was recovered in another vial. Repeated extractions with methanol/chloroform were performed until the supernatant became transparent. The recovered supernatants were then dried under a gentle stream of clean nitrogen gas and redissolved in 1 mL of methanol/acetone (8:2 v/v). The dissolved samples were then passed through a 0.22 µm-pore polytetrafluoroethylene (PTFE) syringe filter and transferred to HPLC vials for analysis.

For the freeze-dried extract (supernatant), 3 mL of methanol was added to 20 mg of the freeze-dried sample in an inactinic vial. The mixture was shaken in an ultrasound bath for 5 min, passed through a 0.22 µm-pore PTFE syringe filter, and transferred to HPLC vials. For the liquid extract, pigments were extracted from the liquid supernatant (for comparison with the dried supernatant) as described by Safi et al. [13].

Pigment analysis was performed by HPLC with a Dionex UltiMate™ 3000 quaternary pump (Thermo Fisher Scientific, Illkirch-Graffenstaden, France) and a C18 column (150 mm × 4.6 mm inner diameter, 5 µm particle size, Agilent, Les Ulis, France). The pigments were eluted in acetone/methanol/water (55:25:20) at a flow rate of 1 mL/min.

2.6. Statistical Analysis

Experiments performed separately for the two techniques for kinetic studies and for the optimization of conditions were performed only once and analyzed at least twice due to the limited amount of biomass available and the number of experiments to be performed.

For a final comparison of HPH and US, experiments were conducted in triplicate, and each sample was analyzed at least twice. All analysis and characterization data are presented as the means ± standard deviations of three experiments. Statistical analysis was performed by one-way ANOVA in Microsoft Excel 2019. The results were reproducible to within ±5%.

3. Results and Discussion

3.1. Ultrasonication (US)

This batch of Tetraselmis suecica had a DW concentration of 20% on reception, and was diluted to 6.6% DW for further processing. The initial biomass had a total protein content of 20.14 ± 0.01% DW, as determined by the Kjeldahl method.

The protein concentration of the extract was greater at higher amplitude (70%), reflecting more intense cell lysis. The protein extraction yield increased with longer treatment times at low amplitude (20%), from 22% at 5 min to 51% at 25 min (Figure 1). These findings are consistent with those of Delran et al. [26], who observed similar trends with the rupture of Tetraselmis suecica cells (10% DW) at a power of 100 W. Progressive changes in the internal structure of microalgae over time facilitate the solubilization and release of proteins. At an amplitude of 70% (equivalent to 350 W), the protein concentration reached a peak of 2.2 mg/mL after 10 min, corresponding to the release of 68% of the protein present. However, the protein concentration decreased beyond this point, possibly because the harsh conditions led to protein denaturation [35]. Better results were obtained for an amplitude of 70% than for an amplitude of 20% at all time points studied, with a faster extraction of larger amounts of protein. These findings indicate that cell lysis is more potent at 70%, facilitating rapid protein release, albeit with a risk of protein denaturation. The increase in protein extraction was attributed to the ultrasound-induced perforation of microalgal cell walls. These pores enhance cell wall permeability, facilitating solvent penetration and protein release [36].

The variation in the fraction of water-soluble proteins released over time followed a polynomial trend for both amplitudes (Figure 2). This trend fit the 20% amplitude data exceptionally well (R² = 0.9971), whereas the fit was slightly less accurate for an amplitude of 70% (R² = 0.9371). The decrease in the protein concentration should correspond to protein denaturation and may disturb the kinetic model for the 70% amplitude. Even if proteins are extracted at this time, the measured concentrations do not correspond to the actual level of protein release. Therefore, the kinetic parameters were also calculated for the initial points before the degradation, and the model was referred to as the “a = 70% modified”.

The protein release fraction was time-dependent and this relationship was characterized by two different models based on the amplitude or power used:

$$\mathrm{X}=-0.6345{\mathsf{\zeta }}^2+1.1281\mathsf{\zeta }+0.0305 \mathrm{for} \mathrm{a}=20\%$$

(10)

$$\mathrm{X}=-1.5902{\mathsf{\zeta }}^2+3.0933\mathsf{\zeta }+0.026 \mathrm{for} \mathrm{a}=70\%$$

(11)

Upon limiting our model to ζ = 0.4 for the 70% amplitude, corresponding to the observed increase in the protein concentration without denaturation, we obtained a well-fitted polynomial trend with R² = 1 (Equation (12)).

$$\mathrm{X}=-3.605{\mathsf{\zeta }}^2+3.0933\mathsf{\zeta }+0.0261 \mathrm{for} \mathrm{a}=70\%$$

(12)

For the determination of the diffusion kinetics of the proteins, we defined the release rate constant, k, which was obtained by the graphical solution of Equation (8). The difference in protein release rates was explained by the kinetic model. Proteins diffused from the cells four times faster at an amplitude of 70% than at an amplitude of 20%, with protein release constants of 0.0005 s⁻¹ and 0.0019 s⁻¹, respectively. However, the fit of the kinetic model was less well suited for the 20% amplitude (R² = 0.9091). When the last concentration point was eliminated from the model, the fitted linear trend became more accurate, and the model was better validated (R² = 0.9804). In this work, we used the value of k calculated from all the data points without model adjustment, as this value represents a more realistic model for the kinetics of protein release. It is clear from Figure 2 that the variation in slope between the two amplitudes reflects the speed of protein extraction, which is greater at 70% than at 20%. This observation indicates that the access of proteins to water is greater at higher amplitudes, facilitating their release. The calculated release rate constant provides a qualitative explanation for the mechanism of protein diffusion. Suarez Garcia et al. [37] explained that aqueous proteins are rapidly released from the cytosol and internal organelles during cell lysis. Functional proteins are released later as the lysis conditions become more severe, due to their less accessible locations, with proteins involved in light capture located in the chloroplasts, for example. The concentration of the extracts at steady state indicated that increasing the severity of treatment led to an increase in the amount of protein that could be extracted. Nevertheless, structural proteins in the cell wall may undergo denaturation during cell disintegration because they are released by osmotic shock during the dilution of the paste.

The total carbohydrate content of this batch of Tetraselmis suecica was 9.1 ± 0.01% DW for samples analyzed in duplicate. This value is within the range of carbohydrate levels (8–57% DW) reported in previous studies on Tetraselmis suecica [13]. The main sugars were galactose, glucose, and mannitol, with traces of fucose. Ribose, arabinose, and mannose were also detected, but at concentrations below the limit of quantification, and xylose was entirely absent. The extraction of carbohydrates for the two amplitudes tested (20% and 70%) showed that 10 min of extraction was sufficient for the release of 35% of carbohydrates into the supernatant for the 20% amplitude and 22% for the 70% amplitude (Figure 3). The treatment at 70% is less efficient than at 20%, and a longer treatment time at 70% amplitude shows a decrease in the carbohydrate concentration due to the modification of the extract structure, leading to lower separation efficiency. The lower yield obtained at an amplitude of 20% after 15 min could also be due to these phenomena. Further investigations are required to confirm these hypotheses and rule out experimental error. We deduce from our findings that sonication for 10 min has an effect almost equivalent to that of 25 min of sonication at the 20% amplitude. At an amplitude of 70%, the carbohydrate concentration decreased after periods of exposure to US exceeding 10 min. We hypothesize that long periods of exposure to high-amplitude US may lead to carbohydrate degradation. These results are consistent with those of Di Caprio Fabrizio et al. [38], who demonstrated that carbohydrate extraction rates for T. obliquus reach a maximum at the start of US exposure, subsequently stabilizing during an equilibrium phase. Furthermore, in accordance with the findings of Delran et al. [26], 5 min of sonication was sufficient to extract a maximum of 60% carbohydrates from Tetraselmis suecica with a power of 200 Watts. A comparison of the two amplitudes, 20% and 70%, revealed that an amplitude of 20% (equivalent to 100 W) resulted in higher carbohydrate release rates than an amplitude of 70% (equivalent to 350 W). We can infer from these findings that an amplitude of 20% is optimal for carbohydrate extraction from Tetraselmis suecica (Figure 3).

The fraction of carbohydrates released X_{carbohydrates} at amplitudes of 20% and 70% over a period of 10 min followed a polynomial trend (R² = 1 for both amplitudes) and was based on three times: 0 (which is obtained from the control treatment), 5 and 10 min. This fraction was linked to the time fraction $\mathsf{\zeta }$ by the following equations:

$${\mathrm{X}}_{\mathrm{carbohydrates}}=0.6474{\mathsf{\zeta }}^2+0.3718\mathsf{\zeta }+0.1776 \mathrm{for} \mathrm{a}=20\%$$

(13)

$${\mathrm{X}}_{\mathrm{carbohydrates}}=-0.4469{\zeta }^2+0.3558\mathsf{\zeta }+0.1776 \mathrm{for} \mathrm{a}=70\%$$

(14)

Based on first-order kinetics, the carbohydrate release rate constant was determined by graphically solving Equation (8), as for proteins. Over the 10 min period considered, carbohydrates diffused at a rate of 0.0006 s⁻¹, which was notably slower than the rate of protein diffusion observed at an amplitude of 20% over this period (0.0008 s⁻¹). This finding suggests that diffusion rates are significantly higher for proteins than for carbohydrates. Suarez Garcia et al. [37] explained that carbohydrates diffuse more slowly due to the solubilization of saccharides from the cell wall and starch granules. Interestingly, at an amplitude of 70%, carbohydrate release was much slower than at 20%, at a rate of 0.0002 s⁻¹. An amplitude of 20% therefore appeared to be better in terms of both the carbohydrate yield and carbohydrate release rate.

We analyzed the following pigments in this batch of Tetraselmis suecica: fucoxanthin, lutein, chlorophyll b, chlorophyll a, and β-carotene. We also compared the pigment concentrations of the aqueous extracts obtained under various US conditions. The variations in the pigment concentration with US treatment time for amplitudes of 20% and 70% are shown in Figure 4 and Figure 5, respectively. The concentration of pigments gradually increased with treatment time at both amplitudes tested. During US treatment, the cells were progressively lysed over time until complete fragmentation occurred, allowing the release of pigments. The pigment concentration reached similar levels for the two amplitudes after 25 min. This observation suggests that the microalgal wall is fragile, with even low-amplitude treatments being sufficient for pigment extraction. Natarajan et al. [24] showed that the US power had a minimal effect on the efficiency of cell lysis in Tetraselmis suecica. The pigment content in the supernatant can be explained by the nature of these hydrophobic biomolecules and their localization in chloroplasts, necessitating major cellular disintegration for their release. It is thought that pores form under these conditions, providing the solvent (water) with access to these biomolecules. However, these lipophilic pigments have a limited solubility in water and their levels were not influenced by the use of a high power. We did not perform a kinetic study of the pigments because the total pigment content of the initial biomass was not analyzed for this batch.

3.2. High-Pressure Homogenization (HPH)

This batch of Tetraselmis suecica had an initial DW content of 15% and was diluted to 5% DW for further processing. The maximum temperatures reached 31 °C, 37 °C, and 43 °C for conditions 1, 2, and 3, respectively, without the need for cooling water. These temperatures are optimal for protein solubilization, facilitating protein extraction, and are sufficiently low to prevent carotenoid oxidation, protein denaturation, and chlorophyll degradation. We analyzed the kinetics of biomolecule diffusion in the aqueous supernatant for proteins, neutral carbohydrates, and pigments (chlorophyll a, chlorophyll b, fucoxanthin, lutein, and β-carotene). With the DW content of all the recovered supernatants maintained at 3.2 ± 0.3%, the extraction rates of these biomolecules can be expressed in terms of concentration (mg/L), concentration per g of dry matter (mg/g of DW), or purity.

Total protein accounted for 30.6 ± 0.07% of the DW of the initial biomass, as determined by the Kjeldahl method (equivalent to 5.15% DW of nitrogen; NTP = 5.95). The high levels of protein present necessitated appropriate fractionation to achieve the best possible yields and purities. We monitored the kinetics of protein diffusion at various centrifugation times, ranging from 5 min to 3 h, by analyzing both extracts centrifuged immediately at t₁ and extracts centrifuged after 3 h, at t₆. Such analyses can shed light on the mechanisms of molecule release and the effects of centrifugation, either facilitating molecule release from the residue or favoring molecule retention within the matrix. We used the Kjeldahl method to assess the variation in the total protein content with the centrifugation lag time for the three HPH extraction conditions.

All the extracts had constant protein contents, maintained at 33.1 ± 1.1% DW (Figure S1 in the Supplementary Materials). The protein concentration therefore remained constant at 7.9 ± 1.7 g/L, regardless of the timing of centrifugation after the end of lysis. Centrifugation therefore had no noticeable effect on the diffusion of proteins from the cytoplasm. The variations in the protein concentration observed at the time points studied can be attributed to differences in the dry weight of the prepared samples, as each sample was prepared separately. However, when the results for all samples were normalized for the same dry weight, the differences were negligible. Once the cells were lysed, all proteins were released rapidly. Similarly, the protein extraction yield remained constant across all three conditions at 64.4 ± 1.5% (Figure 6). This value is the mean extraction yield for the 18 samples, corresponding to six analyses for each of the three sets of conditions. Increasing the number of passes from two to five or increasing the pressure from 300 bars to 600 bars had no discernible impact on protein extraction. Proteins were released immediately, even under milder lysis conditions (P = 300 bars and two passes).

These findings suggest that significant cell lysis occurred under these conditions, releasing 64.4% of the proteins. Similar results were obtained by Rida et al. [39], who extracted 61.9 ± 3.9% of proteins from Tetraselmis suecica using 300 bars of pressure and two passes. Moreover, Tetraselmis suecica appears to have a weak cell wall. These results are consistent with those obtained by Safi et al. [40], who demonstrated that two passes were sufficient to extract 66% of the proteins from Chlorella vulgaris, with the protein yield remaining consistent regardless of the number of passes.

The concentration of water-soluble proteins remained relatively constant across all supernatants at 2 ± 0.3 mg/mL. It was not possible to compare the two analytical methods at this point due to differences in their underlying principles. In the Kjeldahl method, the total Kjeldahl organic nitrogen content was quantified, including molecules not necessarily associated with proteins. In contrast, the Bradford method primarily detects residues of three specific amino acids (arginine, lysine, and histidine).

Total carbohydrate accounted for 11 ± 0.1% of the DW of the initial biomass. A constant carbohydrate extraction yield of 30 ± 2% was obtained across all centrifugation time lags and extraction conditions (Figure 7), suggesting that similar amounts of carbohydrate were released by centrifugation immediately after cell lysis and centrifugation after a time lag of 3 h. Once the cell is disrupted, all the soluble carbohydrates are rapidly released, and centrifugation has no detectable effect on sugar diffusion. Furthermore, the carbohydrate content of the aqueous extract remained constant at 6.6 ± 0.6% DW for all three sets of lysis conditions.

Increasing the number of passes from two to five and the pressure from 300 to 600 bars did not affect carbohydrate extraction. Thus, the conditions used here were not harsh enough to extract carbohydrates efficiently from Tetraselmis suecica. These polymers may act as structural molecules of limited solubility in water or form insoluble complexes with other molecules. These complexes tend to settle in the pellet during separation by centrifugation. Alternatively, these water-soluble complexes may be entrapped within lipophilic emulsions, preventing their passage into the supernatant [41]. These findings are consistent with those of Delran et al. [20], who found that with a single pass at 400 bars, the carbohydrate extraction yield reached 60%, remaining constant thereafter, even if the conditions became more severe.

The pigment analysis was conducted on the liquid supernatant, which was stored at −20 °C for one year before analysis. Figure 8, Figure 9 and Figure 10 show the differences in the pigment concentration in aqueous extracts between conditions 1, 2, and 3, respectively, of cell lysis by HPH. Pigment concentrations clearly remained relatively constant over the six centrifugation lag times tested, with only slight variations. These variations can be attributed to the degradation of pigments in water during the storage of the aqueous samples. Centrifugation at three hours after cell lysis did not result in significantly greater pigment diffusion than centrifugation at five minutes after lysis. Such behavior would typically be expected for lyophilic pigments, which tend to move toward the pellet during centrifugation. The relatively low pigment concentration in the supernatant can be attributed to the probable formation of complexes between certain pigments and water-soluble proteins during the centrifugation phase.

Increasing the number of passes from two to five resulted in higher concentrations of fucoxanthin and lutein, increasing from 0.06 to 0.2 mg/g DW for each pigment, of chlorophyll b, increasing from 0.3 to 1 mg/g DW, of chlorophyll a, increasing from 0.4 to 1.1 mg/g DW, and of β-carotene, increasing from 0.04 to 0.08 mg/g DW. However, increasing the pressure from 300 bars to 600 bars resulted in only a slight increase in pigment concentrations, except for chlorophyll a and β-carotene. The number of passes appeared to have a greater impact than homogenization pressure. This increase in pigment concentrations suggests that the chloroplasts were disrupted, allowing water unhindered access to the interthylakoid space, where certain pigments are located [13]. The extraction of pigments can be explained as a function of their cellular location. Chlorophyll a is typically found in photosystem I (PSI), which is situated on the surface of the thylakoid in contact with the stroma, making it more accessible to the solvent than chlorophyll b, which is typically located in photosystem II (PSII). This accounts for the concentration of chlorophyll a being higher than that of chlorophyll b.

3.3. Comparison of HPH and US

The preceding section addressed the comparison of various HPH conditions and their effects on biomolecule extraction. The optimal conditions for HPH were determined to be two passes at a pressure of 300 bars, with no improvement in extraction yield observed with further increases in the pass number and pressure. We then showed that an amplitude of 20% and a treatment time of 25 min were the ideal conditions for US, striking a balance between protein and carbohydrate yields while minimizing energy consumption. We then performed tests on a single batch of cells under the optimal conditions defined above to compare HPH and US in a meaningful manner and to identify the most efficient extraction method with the highest yield for use in the subsequent biorefining of Tetraselmis suecica.

3.3.1. Dry Matter

The batch of Tetraselmis suecica used had an initial DW content of 15% and was diluted to 5.65 ± 0.05% DW for further processing. We first compared the dry matter extraction yield between the two disruption techniques (HPH and US). HPH extracted 52.1 ± 1.7% of dry matter, whereas US extracted 46.4 ± 1.5% of dry matter. Both methods were more effective than the control, which extracted only 24.1 ± 0.3% of the dry matter (Figure 11). Dry matter extraction is crucial for microalgal valorization and for assessments of the efficiency of cell lysis. However, a comprehensive comparison between the two techniques also requires the characterization of the extract and a determination of the content of the biomolecules of interest within it. Many studies have focused on the extraction yield without considering the metabolite content. Suarez Garcia et al. [37] emphasized the importance of using both the yield and content to guide subsequent purification steps and to enhance the technical classification of these products.

The amount of dry matter recovered in the supernatant corresponds to the biomolecules extracted, as these molecules tend to migrate to the aqueous phase during centrifugation. We can therefore conclude that HPH is a more effective cell lysis technique than US, resulting in higher extraction rates. These results contrast with those of Halim et al. [23], who found that US and HPH were similarly effective for the lysis of Tetraselmis suecica cells but that US gave better results for lipid extraction. They suggested that the extracted lipids might be degraded under the HPH conditions used. In contrast, Zhang et al. [19] compared HPH and US for the lysis of P. kessleri and concluded that US damaged the cells less effectively, whereas HPH led to the generation of large amounts of cell debris.

3.3.2. Minerals

The biomass received had a high mineral content, at 29.6 ± 0.6% DW. The primary goal of microalgal fractionation is to extract valuable molecules with important industrial applications, such as pigments, proteins, and carbohydrates. The control extract (obtained by lysis in water) had the highest mineral content, at 71.2 ± 1.02% DW (Figure 12), due to the hydrophilic nature and high solubility of these molecules in water. The US extract had a higher mineral content (43.8 ± 1.8% DW) than the HPH extract (35.5 ± 0.8% DW). These findings suggest that HPH is more effective for the extraction of intracellular biomolecules than US, leading to the release of fewer minerals during subsequent aqueous washing. This is an advantage of HPH over US, as HPH yields an extract with fewer minerals and higher concentrations of other compounds of interest.

3.3.3. Proteins

The total protein content of the initial biomass was 23.1 ± 0.3% DW, as analyzed by the Kjeldahl method. The protein concentration was determined by the Bradford method with BSA as the standard for the comparison of biomass treatments (Figure 13) and with the Kjeldahl method for the yield calculation (Figure 14). Both these methods showed that HPH cell disruption resulted in a higher protein concentration than US treatment. HPH extracted a higher proportion of proteins (61.9 ± 3.9%) than US (44.8 ± 1.3%). In addition, the protein content of the HPH extract (26.1 ± 0.3% DW) was higher than those of the US extract (22.6 ± 0.3% DW) and the control extract (4.1 ± 0.4% DW). These findings highlight the superiority of HPH over US for protein extraction. Importantly, the conditions applied (two passes at 300 bars) were not very severe and were energy-efficient. Kinetic studies in the initial phase revealed that protein extraction with HPH was rapid and instantaneous, whereas protein extraction by US treatment required 25 min. These results are consistent with the findings of Safi et al. [29], who reported higher protein extraction yields with HPH than with US for various microalgae. HPH has also been shown to be more effective at extracting proteins from P. kessleri while consuming less energy, whereas US conditions are thought to cause protein degradation [19]. Furthermore, two separate studies conducted by Delran et al. [20,26] on the rupture of Tetraselmis suecica cells indicated that HPH at 400 bars allowed the extraction of 80% of the proteins, whereas US at 200 W power resulted in the extraction of 88% of the proteins after 60 min of treatment. The pressure range of 200–400 bars used in HPH corresponds to the threshold for cell disruption, with rupture of the membrane allowing the solvent access to proteins. Delran and coworkers also reported that the use of a power of 200 W resulted in changes in the shape of Tetraselmis suecica cells after 30 min of treatment, consistent with lower rates of cell lysis for US treatment than for HPH. This difference in protein extraction yields between the two techniques can be attributed to the extent of cell disruption. Based on the content, extraction yield, concentration, and protein release rate kinetics of the two techniques, HPH appears to be a more efficient method for the subsequent fractionation process.

3.3.4. Carbohydrates

The total carbohydrate content of the initial biomass was 7.3 ± 0.7% DW. The sugars identified in the initial biomass were mannitol, fucose, arabinose, galactose, glucose, mannose, and ribose. The monosaccharides identified are consistent with those detected by Pereira et al. [42]. Kermanshahi-Pour et al. [43] demonstrated that the structural monosaccharides present in the cell wall of Tetraselmis suecica are 3-deoxy-d-manno-oct-2-ulosonic acid (KDO), 3-deoxy-lyxo-2-heptulosaric acid (DHA), galacturonic acid, and galactose. The carbohydrate extraction yield did not differ significantly between the two techniques used (p > 0.05), with a value of 35% obtained for both of them (Figure 15). US (5.4 ± 0.7% DW) resulted in a slightly higher carbohydrate content than HPH (4.8 ± 0.6% DW), but the difference was almost negligible. These findings suggest that both methods are suitable for carbohydrate extraction. However, the kinetics of carbohydrate extraction remain an important factor for comparison. We found that carbohydrates were extracted significantly more rapidly with HPH than with US, for which 25 min of treatment were required. This is a crucial consideration, with HPH identified here as the superior technique for extracting carbohydrates from Tetraselmis suecica. In a study by Delran et al. [20,26], the carbohydrate extraction yields from Tetraselmis suecica were investigated in two separate experiments for HPH and US. US resulted in the extraction of 70% of carbohydrates at a power of 100–200 W after 60 min of treatment, whereas HPH released 60% of carbohydrates at a pressure of 400 bars. Zhang et al. [19] showed that carbohydrate extraction is correlated with energy consumption. For example, using US at a power of 400 W for 30 min allowed the extraction of 45% of carbohydrates from P. Kessleri with an energy consumption of 56 kJ/g DW, whereas HPH at a pressure of 400 bars enabled the extraction of 20% of carbohydrates with a consumption of 16 kJ/g DW.

3.3.5. Uronic Acids

Total uronic acids in Tetraselmis suecica accounted for 17.1 ± 0.8% DW of the initial biomass. Uronic acids were extracted more efficiently by US, with an extraction yield of 24 ± 4% and a uronic acid content of 8.9 ± 1.8% DW. In comparison, HPH allowed the extraction of 20 ± 6% with a uronic acid content of 6.4 ± 2% DW. Both of these techniques outperformed the control, which released only 5% of uronic acids, with a content of 2.9 ± 0.2% DW (Figure 16). The difference in uronic acid extraction rates between HPH and US can be attributed to the impact of these molecules on cell structures. Galacturonic acid is located primarily in the cell wall in Tetraselmis suecica. US is a less severe cell disruption method than HPH and is not thought to destroy the cell wall of Tetraselmis suecica, instead creating pores via which uronic acids are released. Conversely, HPH is a more severe technique that fragments the cell wall, leading to uronic acid degradation. In terms of kinetics, HPH is considered superior to US because it leads to an immediate release of molecules, whereas a treatment time of 25 min is required for US.

3.3.6. Pigments

The pigments detected and analyzed in Tetraselmis suecica were fucoxanthin, lutein, chlorophyll b, chlorophyll a, and β-carotene. The total pigment content of the initial biomass was 0.3 ± 0.07% DW. HPH was more efficient at pigment extraction, with an extraction yield of 25.7 ± 3.4%, than US, for which the extraction yield was much lower, at 11.2 ± 1.3%. This difference highlights the ability of HPH to facilitate the extraction of intracellular molecules more effectively than US, leading to greater cell lysis. The higher pigment content of the HPH extract (0.14 ± 0.02% DW) than of the US extract (0.07 ± 0.007% DW) suggests major alterations to the chloroplasts, releasing pigments from the thylakoids. Conversely, the low-intensity conditions used for US resulted in milder damage to the cells, impeding the passage of water across the phospholipid membrane of the thylakoid. The low extraction yields for pigments may reflect their hydrophobic nature, causing them to migrate toward the pellet. However, their presence in the aqueous phase was explained by Safi et al. [13], who suggested that pigments adsorb onto fine cell debris that does not settle in the pellet during centrifugation or are present in emulsions or attached to phospholipids. Similar findings were reported by Zhang et al. [19], who demonstrated that the extraction of pigments from P. kessleri was more effective with HPH than with US. Comparisons of the extraction yield and pigment content in Tetraselmis suecica extracts suggest that HPH is more effective for pigment extraction than US. The kinetics of pigment release were also better for HPH, which allowed faster extraction than US.

4. Conclusions

This study sheds light on the influence of cell disruption techniques and conditions on the extraction of biomolecules and the kinetics of their release behavior. A screening of extraction conditions based on mechanical cell lysis treatments revealed that both HPH and US were viable methods for the fractionation of Tetraselmis suecica. We evaluated and compared these rupture methods on the basis of the extraction yield, biomolecule content of the extract, and biomolecule release kinetics. Based on the parameters investigated, HPH is better for protein and pigment extraction, whereas US can release larger amounts of uronic acid. Interestingly, the two methods had similar extraction yields for carbohydrates. The release kinetics for all biomolecules showed that HPH was superior to US, resulting in the rapid release of metabolites. This study focused on the kinetics of molecule extraction, but additional studies are required to understand the morphological changes in cells over time. Such studies would provide detailed explanations for the phenomenon of biomolecule diffusion observed during cell disruption. The use of sequential cell disruption methods might increase the extraction yields for both the proteins and carbohydrates of Tetraselmis suecica. The initial focus could be on carbohydrate extraction, with a shift towards protein recovery in subsequent steps. Further studies are required for process optimization, and a techno-economic assessment is required to assess the feasibility of this approach. Our findings indicate that proteins are extracted alongside polysaccharides and other molecules, resulting in a lower protein content in the final extract. Additional separation techniques are required to purify target molecules by eliminating the molecules extracted with them.