Unveiling potential of promising filamentous microalga Klebsormidium cf. nitens: Shear stress resilience and carotenoid-fatty acid dynamics in tubular photobioreactor

Bioresource Technology 407 (2024) 131147 Available online 21 July 2024 0960-8524/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Unveiling potential of promising filamentous microalga Klebsormidium cf. nitens: Shear stress resilience and carotenoid-fatty acid dynamics in tubular photobioreactor F.J. Segura-Morales a , A. Molina-Miras a , M.C. Cer´ on-García a,b , A. S´ anchez-Mir´ on a,b , S. Seoane c,d , A. Contreras-G´ omez a,b,* , F. García-Camacho a,b a Department of Chemical Engineering, University of Almería, 04120 Almería, Spain b Research Centre CIAIMBITAL, University of Almería, 04120 Almería, Spain c Department of Plant Biology and Ecology, 48940 Leioa, Spain d Technology and Research Centre for Experimental Marine Biology and Biotechnology, University of the Basque Country (UPV/EHU), 48620 Plentzia, Spain HIGHLIGHTS GRAPHICAL ABSTRACT •K. cf. nitens grows consistently at high shear stress (6600 mPa) •Fragmentation of large filaments favours vegetative multiplication of K. cf nitens. •K. cf. nitens grows in a tubular photobioreactor at high fluid velocity (70 cm s −1 ) •Xanthophyll cycle is activated to protect cell photosynthetic system. •Linoleic acid makes over 50 % of total fatty acids. ARTICLE INFO Keywords: Streptophyta Pump-driven Semicontinuous Xanthophyll cycle Linoleic acid ABSTRACT In this study, the effects of shear stress and different culture media on the growth of the filamentous microalga Klebsormidium cf. nitens were studied. The microalga’s growth, carotenoids and fatty acids were further evaluated in a pump-driven tubular photobioreactor. The results show that this microalga had the ability to withstand high shear stress and the adaptability to grow in a culture medium that lacks certain trace elements. K. cf. nitens grew consistently in the tubular photobioreactor at different average light intensities although it did not grow well in a tall bubble column. The carotenoid analysis revealed that the xanthophyll cycle was activated to protect the cell photosynthetic system. The fatty acids increased with irradiance, with linoleic acid (C18:2n6) making up over 50 % of the total fatty acids. This study supports the potential of employing pump-driven tubular photobioreactors to produce the filamentous microalga K. cf nitens at the large scale. * Corresponding author at: Department of Chemical Engineering, University of Almería, 04120 Almería, Spain. E-mail addresses: [email protected] (F.J. Segura-Morales), [email protected] (A. Molina-Miras), [email protected] (M.C. Cer´ on-García), [email protected] (A. S´ anchezMir´ on), [email protected] (S. Seoane), [email protected] (A. Contreras-G´ omez), [email protected] (F. García-Camacho). Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech https://doi.org/10.1016/j.biortech.2024.131147 Received 12 March 2024; Received in revised form 28 June 2024; Accepted 20 July 2024 Bioresource Technology 407 (2024) 131147 2 1. Introduction Over the last few years, algae have become increasingly relevant in biotechnological research due to their promising potential as a source for countless products, from chemicals such as carotenoids and polyunsaturated fatty acids, to biofertilizers or biofuels (Borowitzka, 2015). However, the development of commercial processes has not yet materialised, largely because of the high production and harvesting costs (Guo et al., 2014), with the exploitation of microalgae being mainly limited to the use of a few unicellular species in small high-value market niches, such as certain cosmetic or nutraceutical products. (Gellenbeck, 2012; Xu et al., 2021). In this context, finding new species may be a key factor in lowering the elevated costs inherent in current microalgae production systems (Xu et al., 2021). Filamentous eukaryotic microalgae have attracted great interest in this regard due to their high growth rates, resistance to predators, and ease of harvesting and dewatering, so potentially providing significant advantages over unicellular microalgae when it comes to developing commercial processes that contribute to the global bioeconomy (Long et al., 2022). Klebsormidium (Archaeplastida, Phylum Streptophyta, Class Klebsormidiophyceae) is a cosmopolitan green, non-branched filamentous genus widely distributed across different terrestrial and freshwater habitats. It has garnered considerable interest in recent years for several compelling reasons: from a physiological and ecological standpoint, Klebsormidium species exhibit high photophysiological plasticity under variable irradiance (Karsten et al., 2013; Míguez et al., 2020) and are relevant members of the communities found in biological soil crusts (Herburger et al., 2016) where they act as important primary producers, contributing to carbon and nitrogen cycling, water retention, and soil stabilization (Elbert et al., 2012). The intriguing capacity of Klebsormidium to adapt to the extreme environmental conditions of their habitats has also received much attention (Holzinger and Karsten, 2013; Chatelain et al., 2022). Furthermore, the four Klebsormidiophyceae genera (Entransia,Hormidiella,Interfilum, and Klebsormidium), contain genes unique to terrestrial plants, making them remarkable in an evolutionary context as potential ancestors of land plants (Lewis and McCourt, 2004). From a commercial standpoint, Klebsormidium has been proposed as a candidate for lipid and fatty acid production (Xu et al., 2021) as well as for compounds exhibiting different activity such as mycosporine-like amino acids (Kitzing and Karsten, 2015), phytohormones (StoynevaG¨ atner et al., 2019) or anti-inflammatory compounds (Qiu et al., 2020). Klebsormidium has also been reported as efficient in treating wastewater to eliminate phosphorus and nitrogen (Valchev et al., 2021; Umetani et al., 2024) and is particularly valuable in aquaculture, where it is used as an aquatic feed additive due to its favourable composition and high nutritional value (Fang et al., 2022). Additionally, Klebsormidium offers a promising solution in supplying the increasing demand for plant-based protein sources within the food industry (Brickel et al., 2018). Klebsormidium has recently undergone safety evaluations that have demonstrated its potential as a safe and viable food ingredient, thus further supporting its exploration as a novel food source (Brickel et al., 2018). Its combined phycological and nutritional properties make it an environmentally friendly and sustainable source of food and aquatic-feed supplements since it can be grown using significantly less land, water, and resources than other traditional animal-based protein sources. Comprehensive studies evaluating the potential of Klebsormidium for large-scale culture production are scarce. Only one recent study has explored this possibility using the strain Klebsormidium sp. LGX80 (Xu et al., 2021). Nonetheless, the authors acknowledge that the shear stress present in closed airlift-driven tubular photobioreactors may contribute to reduced biomass production compared to laboratory-scale bubblecolumn cultures (Xu et al., 2021). Other studies have highlighted the great morphological diversity of Klebsormidium species and their varied responses to environmental conditions (Mikhailyuk et al., 2014). This diversity may introduce significant intraand interspecific variability in the cell response, thus requiring further investigation. Furthermore, the shear stress profile differs depending on whether the photobioreactor is driven by an airlift pump or a centrifugal pump, particularly in the degasser and culture pumping zones. The objective of this study, therefore, is to expand the portfolio of Klebsormidium species that have commercial potential by assessing the growth performance of the K. cf. nitens strain in a pilot-scale tubular photobioreactor driven by a centrifugal pump. Initially, a brief laboratory-scale study (1 L culture flasks) was carried out to evaluate its sensitivity to hydrodynamic forces and its growth potential in different culture media. Subsequently, mass cultivation assays were conducted in a 12 L vertical-fence tubular photobioreactor illuminated with an LEDbased system and driven by a centrifugal pump. Sequential batch and semicontinuous culture modes were tested along with different light- –dark cycles. The biomass harvested from the tubular photobioreactor was analysed to evaluate the carotenoid and fatty acid content. This study aims to provide a preliminary assessment of the potential for using tubular photobioreactors in the large-scale cultivation and commercial exploitation of K. cf. nitens. 2. Materials and methods 2.1. Microalga strain and maintenance The freshwater microalga used in this study was Klebsormidium cf. nitens, which is adapted to brackish water. It was isolated in May 2007 from a water sample collected in a mesohaline area of the Bilbao Estuary (Bilbao, Spain) and deposited in the Basque Microalgae Culture Collection as BMCC149. The identification was based on its morphology under light microscopy and the internal transcribed spacer (ITS) sequence (Primers ITSA and ITSB). The ITS sequence is deposited in GenBank under the accession number OQ451588. Comparing it to the GenBank sequence, the greatest similarity was with K. nitens; however, given the uncertain taxonomy of some species within this genus (Mikhailyuk et al. 2015), identifying it as Klebsormidium cf. nitens is recommended. The inocula were maintained in static T-flasks at 21 ±1◦C under a 12:12 h light–dark cycle. Illumination was provided by an overhead LED panel and the irradiance on the surface of the culture flasks was 250 µE m −2 s −1 . Modified f/2 medium (Andersen et al., 2005) was used for inocula maintenance with an augmented phosphate concentration (to reach a N:P=5), prepared using filter-sterilized brackish water at 15 g/L salt (see supplementary material). The salinity of the brackish water was adjusted by mixing distilled water and Mediterranean seawater in a 2:1 ratio. 2.2. Computational fluid dynamics simulations Traditionally, spherical flasks have routinely been used to carry out all types of microalgal bench-scale research for both non-aerated and aerated cultures. Despite their great practical value, no attention has been paid to the engineering aspect of these bench-scale photobioreactors in terms of how aeration modifies the hydrodynamics, and how this potentially affects microalgal growth. As a first step in this study, the computational fluid dynamics (CFD) approach was used to quantitatively characterise the flow field and to map the energy dissipation rate (EDR) and shear stress inside a 1000 mL spherical flask containing 750 mL water. Furthermore, different gas flow rates were studied to see how the hydrodynamics influenced K. cf. nitens growth in this culture system. The time-dependent simulations were performed on an HP Z840 workstation consisting of two Intel®Xeon®processors, a ES-2670 v3 @ 2.3 GHz CPU and 128 GB of RAM using the commercial Ansys Fluent® v2020R1 suite (www.ansys.com). The realizable kε turbulence model was used to obtain the average values of the variables over time, and the VOF two-phase model was used to describe the liquid–gas interactions. F.J. Segura-Morales et al. Bioresource Technology 407 (2024) 131147 3 The 1000 mL flasks were modelled using the Ansys Modeler®. Based on this, an automatic tetrahedral mesh was generated in the Ansys integrated meshing module. To capture the bubbles’gas–liquid interface in greater detail and the velocity gradients during their formation and surface disengagement, a more refined mesh (with elements having a maximum size of 100 µm) was generated in the area where the air was introduced and where the bubbles ascended. The mesh generated in this way consisted of 3.5 ×10 6 elements. To improve the convergence of the simulations, the tetrahedral mesh was polyhedrally transformed in Fluent®. Consequently, elements with a minimum orthogonal quality of 0.7 were obtained. The fluids used were brackish water (density =1015 kg m −3 , viscosity =1013 µPa s) and air (density =1225 g m −3 , viscosity =17.9 µPa s), being the continuous and dispersed phases, respectively, in all the simulations. The Grace model was used to estimate the drag force between phases using a surface tension =71.8 mN m −1 . For the air inlet and outlet conditions, a volumetric fraction of 1 and a constant pressure of 1 atm were defined. Simulations were performed using a pressurebased model. Gravity in the negative z-direction was included and the pressure at the outlet was used as the reference. The other models used were: PISO for the pressure–velocity coupling; the Least Square Cell Base for Gradient (LSCBG) scheme for spatial discretization; the modified HRIC for the volume fraction; the second-order upwind discretization for the momentum, dissipation rate, and turbulent kinetic energy; and the first-order implicit for transient formulation. Adaptive time steps with a global Courant number of 1, a minimum step size of 1 µs and a maximum of 30 iteration/time step were used. The residual for the absolute convergence criteria were fixed at five orders of magnitude. Data were time averaged with a sampling interval of 10 ms for 5 s. To validate the simulation, the equivalent diameter of the bubbles, obtained by measuring their cross-sectional area, was approximated with the ImageJ image analysis program. 2.3. Assays to determine the shear sensitivity of K. cf. nitens To assess the sensitivity of K. cf. nitens to the hydrodynamic forces associated with turbulence and bubbling, the microalga was grown in 1000 mL flat-bottom spherical flasks in batch mode at different gas flow rates. The working volume was 750 mL. Cultures were grown in a thermostatic chamber where the temperature was maintained at 21 ± 1◦C. The irradiance was provided by an overhead LED panel. The incident irradiance (I 0 ) on the culture flask surface was 400 μ E m −2 s −1 under a 12:12 h light–dark cycle. The culture medium was the same as that used to grow and maintain the inocula, as described in Section 2.1. The initial pH was adjusted to 8.5 with HCl (0.1 mol/L) or NaOH (0.1 mol/L) and was allowed to evolve freely. The aeration flow rates tested were 100, 250, 500, 750, 1000, 1500, 2000 and 3000 mL min −1 . All the experiments were duplicated. 2.4. Influence of different culture media on growth The influence of five different culture media on the growth of K. cf. nitens was studied. The culture media were prepared using the same brackish water as that used to prepare the culture medium used to grow and maintain the inocula, as described in Section 2.1. The culture media were f/2 (N:P=5), f/2 ×6 (N:P=5), L1-Se ×6 (N:P=24), K-PO 4 ×6 (N: P=88), and 3NBBM+V (N:P=4) (Andersen et al., 2005). The culture media f/2 ×6, L1-Se ×6, and K-PO4 ×6 were formulated so they had the same nitrate concentration as 3NBBM+V (see supplementary material). The culture system was the same as that used to study the influence of hydrodynamic forces, as described above: namely, 1000 mL spherical culture flasks with a 750 mL culture volume, a temperature of 21 ±1◦C, an I 0 of 400 μ E m −2 s −1 under a 12:12 h light–dark cycle, and an initial pH adjusted to 8.5 with HCl (0.1 mol/L) or NaOH (0.1 mol/L). The gas flow rate providing the best growth results in the previous assay, 1000 mL min −1 , was used in all the cultures. The cultures were started in batch mode with inocula from the mid-exponential growth phase acclimated to the different culture media. To ensure comparability between the different media, the initial biomass concentration was the same (40 mg L -1 ) in all the cultures. From day 10, the cultures were operated in semicontinuous mode for 18 days (28 days of total culture time). From day 10 to day 25, 100 mL of culture was removed, and 100 mL of fresh culture medium was added every 3 days. All the cultures were duplicated. 2.5. Cultivation in the photobioreactors To assess the potential of K. cf. nitens to grow in a larger-scale system, two popular photobioreactor configurations were used −tubular and tall bubble-column photobioreactors (Fig. 1). The first consisted of a 12L vertical-fence tubular photobioreactor (T-PBR) comprising a light collector (22 m-long loops of Plexiglass tube, with an internal diameter of 19 mm and an external diameter of 25 mm, connected by U-bends and PVC couplers, having a total length of 30 m) connected to a bubblecolumn-type degasser where the photosynthetic O 2 was degassed and CO 2 supplied to the culture. To circulate the microalgal culture, the photobioreactor was equipped with a variable frequency magneticallycoupled centrifugal pump (Xylem Flojet NDP 35/3, Hoddesdon, UK) providing a pumping speed of 700 L h −1 . The shear stress profile in the photobioreactor at operational pumping speed and the residence time in each section were calculated as described previously (Chisti, 2009; Macías-de la Rosa et al., 2023). The bubble-column photobioreactor (BC-PBR) comprised a clear plastic (polymethyl methacrylate) vertical tube (internal diameter 84 mm; height 200 cm) with a culture volume of 10 L (a culture height of 180 cm and an aspect ratio, AR=21). The culture was agitated by injecting filtered air through a single-nozzle sparger (with a nozzle diameter of 12 mm). For cultivation, the photobioreactors were placed in a thermostatic chamber where the temperature was maintained at 21 ±1◦C. The pH was controlled at 8.5 by automatically injecting pure CO 2 from a highpressure cylinder on demand. CO 2 was injected through a solenoid valve in the air flow continuously supplied to the photobioreactor into the degasser of the T-PBR and into the bottom of the BC-PBR. When the pH exceeded the set point due to the metabolism of the microalga, the valve opened to inject CO2, which lowered the pH of the culture. Once the pH reached the desired value, the CO 2 injection automatically stopped. The culture medium providing the best growth results in the previous assays (f/2 ×6; as described in Section 2.4) was used for cultivation. The experiment started in batch culture phase (set 1) inoculated with algal cells in the late exponential growth phase. Once the culture entered the stationary phase, semicontinuous operation was started. For semicontinuous operation, 3000 mL and 2500 mL of culture volume were removed from the T-PBR and BC-PBR, respectively. Fig. 1. Diagram showing the different bubble column and tubular photobioreactors parts and their interconnections. High of the bubble column 2 m. Total length of tubular photobioreactor light collector 30 m. F.J. Segura-Morales et al. Bioresource Technology 407 (2024) 131147 4 Immediately after, an equal volume of fresh medium was added. This was repeated twice. Each time, the nitrate and phosphate concentrations in the medium were measured and fresh medium was supplemented with nitrate and phosphate stock solutions to achieve final concentrations of these two nutrients in the culture similar to the values found in the f/2 ×6 medium. The other nutrients were added in equivalent quantities to achieve the f/2 ×6 medium concentration. LED strips provided illumination to both PBRs. In the T-PBR, the LED strips were placed vertically in the light collector. In the BC-PBR, they were placed in the column in a helical configuration. The LED strips provided a photosynthetically active irradiance of 600 μ E m −2 s −1 , as measured in the centre of the tubes. A light-dark cycle of 12:12 h in the batch culture (set 1), and 18:6h and 24:0h in the first (set 2) and second (set 3) semicontinuous cultures, respectively, were tested. Therefore, the daily average I 0 was 300 μ E m −2 s −1 , 450 μ E m −2 s −1 , and 600 μ E m −2 s −1 , respectively. 2.6. Analytical methods The biomass dry weight was determined periodically in duplicate from samples taken throughout the culture. The total nitrate and phosphate in the supernatants were measured using the 4500-N and 4500-P spectrophotometric methods. The maximum photochemical yield of photosystem II (F V /F M ) was determined using a pulse amplitude modulation (PAM) chlorophyll fluorometer (Mini-PAM-2500; Heinz Walz GmbH, Effeltrich, Germany). The carotenoid content and profile were determined using an HPLC photodiode array detector whereas the fatty acid content and profile were determined by direct transesterification and gas chromatography (Molina-Miras et al., 2018), and total lipids as described by Kochert (1978). 2.7. Statistical analysis The experimental values from the flask cultures were presented as the average values of two independent experiments (i.e. biological replicates) and their standard deviation. Experiments in the T-PBR were based on a single-system design (SSD), meaning they were conducted without biological replicates but with two technical replicates (i.e., two samples). The structure of SSD makes it useful for researching interventions in single-PBR systems, which are typical in industrial PBRs. If an intervention with a PBR is effective, it should be possible to observe a change in status from the period prior to the intervention to the period during and after the intervention. The SSD consisted of three components: (a) measurements over time, (b) a baseline phase (set 1), and (c) a treatment phase (set 2 and set 3). Significant difference analysis was performed with a one-way analysis of variance (ANOVA) test. Statistically significant differences in the mean response between factors were fixed at a 5 % significance level threshold (p<0.05). Fisher’s least significant difference (LSD) procedure was the method used to discriminate between the means at the 95 % confidence level. Statistical Fig. 2. Gas holdup (a), fluid velocity (b), local energy dissipation rate (c), and shear stress (d), averaged over 10 s in an x-y plane in 1000 mL spherical culture flasks with 750 mL of fluid phase. F.J. Segura-Morales et al. Bioresource Technology 407 (2024) 131147 5 data analysis was performed using Statgraphics Centurion 19 (version v.19.5.01) statistical software (2023, Statpoint Technologies, Inc., Warrenton, VA). 3. Results and discussion 3.1. Hydrodynamics, growth, and shear sensitivity of K. cf. nitens Fig. 2 shows the gas holdup, liquid velocity, energy dissipation rate (EDR), and shear stress averaged over 10 s in an x-y plane for the representative gas flow rates used in this study. In this system, which resembles a short bubble column with no moving parts, the gas leaving the sparger is the only source of energy. It is in this small region where momentum transfer between the gas and liquid phases occurs, determining the energy dissipation rate and turbulence. In this kind of system, bubble populations critically influence gas holdup and the bubble and liquid velocities, ultimately causing a significant impact that alters the hydrodynamics (Kantarci et al., 2005). The presence of gas in the liquid phase induces pressure variations resulting in intense liquid acceleration. The rising bubbles entrain liquid alongside them and, because of continuity, a sinking recirculating liquid-phase flow is established. Visual observations confirmed a bubbly flow regime at low gas flow rates with small spherical uniform bubbles and with no bubble coalescence or break-up. In contrast, at the highest gas flow rates, a churn-turbulent flow regime was observed where there were large, nonspherical bubbles due to the high gas throughputs −this produced a holdup profile with maximum values in the region above the sparger (Fig. 2a). The velocity distribution indicates that the liquid phase moves up with the rising bubbles, with maximum velocity values and sharper velocity gradients reached in the vicinity of the gas plume. As one moves away from the gas plume, there is a much lower and homogeneous velocity distribution with a smoother, circular motion prevailing along the flask wall (Fig. 2b). Clearly, bubble-induced turbulence is dominant, and the maximum EDR values (Fig. 2c) and shear stress values (Fig. 2d) are found in the region above the nozzle, as with gas holdup. Table 1 presents the different values of the CFD-obtained EDR and shear stress for all the gas flow rates used in this work. As the table shows, the maximum EDR and shear stress values in the region above the sparger are three orders of magnitude greater than the average EDR and shear stress values in the liquid as a whole. As can be deduced from Table 1, the gas flow rate is a critical variable influencing the environment for microalgal growth. The inlet gas must provide adequate mixing to achieve uniform distribution of the light, temperature, and nutrient concentration. However, as previously reported, the hydrodynamic forces associated with gas sparging can significantly impact the growth of particularly sensitive microalgae (L´ opez-Rosales et al., 2019). In this scenario, when K. cf. nitens was grown at different gas flow rates (Q) in the spherical flasks, the effect of Qon the growth rate in the exponential phase (µ max ) and on biomass productivity at the end of the culture (P b ) was statistically significant (F-ratio =21.8, p<0.05 and Fratio =45.8, p<0.05, respectively).The mean µ max and P b values clearly depended on the gas flow rate, as shown in Fig. 3. Post-hoc multiple range tests revealed 20 and 25 pairs of means with statistically significant differences for µ max and P b , respectively. Fig. 3 shows that lower Q values (<500 mL min −1 ) are associated with reduced EDR and shear rates, hampering the mixing efficiency. Conversely, increasing Qenhances mixing, improving mass transfer, CO 2 availability, and light for cells, thus increasing the growth rate and biomass production. The positive correlation of Qwith µ max and P b was evident, peaking at 1000 mL min −1 . This enhancement was substantial, with a more than two-fold increase in µ max (from ~ 0.2 to 0.4 day −1 ), and a more than eight-fold increase in P b (from 3 to 25 mg L -1 day −1 ) relative to nonagitated and non-aerated conditions (Q=0 L min −1 ). Beyond the 1000 mL min −1 threshold (equivalent to 1.3 vvmin), both µ max and P b remain constant, irrespective of further increases in Q. This pattern strongly suggests the existence of a mass transfer limitation at low energy dissipation and a kinetic limitation arising once the system attains its maximum mass transfer capacity (L´ opez-Rosales et al., 2019). This plateau for µ max and P b implies that, above 1000 mL min −1 , the growth rate becomes limited, either by the diffusion transport rate of the CO 2 contained in the bubbled air or the available irradiance (or both). In addition, a stoichiometric limitation is also observed, where the increase in biomass productivity at the end of the culture is controlled by the availability of nutrients dissolved in the culture medium. Fig. 3 highlights a distinctive characteristic of K. cf. nitens −unlike other filamentous microalgae, it can tolerate hydrodynamic stress even at high gas flow rates. Remarkably, no discernible decrease in either the growth rate or biomass productivity is observed even at gas flow rates as high as 3000 mL min −1 (4 vvmin); this is in sharp contrast to well documented examples of microalgal cell damage in sparged photobioreactors (L´ opezRosales et al., 2019; Suzuki et al., 1995). Although the molecular mechanics in microalgae that convert hydrodynamic forces into cellular signals have yet to be identified (L´ opez-Rosales et al., 2019), there is evidence that the action of the fluid forces induces gene expression regulation, thus producing crucial changes in cell physiology, including oxidative stress. This oxidative stress, in turn, causes cytoplasmatic and thylakoid membrane rigidification, leading to decreased membrane fluidity (Los et al., 2013). It has been postulated that this decrease in cytoplasmatic membrane fluidity, which is driven by fluid forces, increases the cells’mechanical resistance (see L´ opez-Rosales et al., 2019). Rigidification of the thylakoid membrane undoubtedly disturbs every aspect of the photosynthetic machinery, especially affecting photosystem II repair, its most vulnerable component (Nishiyama et al., 2004), and thus stimulating photoinhibition (Los et al., 2013). Filamentous microalgae present a unique challenge. Various studies have reported that relatively low shear stress thresholds, exceeding 300 mPa, cause filament breakage, leading to impaired metabolism and negatively affecting cell growth (Michiel et al., 2016; Wang and Lan, 2018). As a result, filamentous microalgae typically exhibit heightened shear sensitivity compared to unicellular microalgae, having reduced cell viability following filament destruction and disintegration (Michiel et al., 2016; Wang and Lan, 2018). However, the present study Table 1 Different CFD-obtained EDR and shear stress for all the gas flow rates. Gas flow rate (mL min −1 ) Whole average EDR (mW kg −1 ) Above the nozzle average EDR (mW kg −1 ) Maximum EDR (mW kg −1 ) Whole average shear stress (mPa) Above the nozzle average shear stress (mPa) Maximum shear stress (mPa) 100 1 11 1615 1 12 2658 250 2 19 4121 2 37 6149 500 4 34 8194 4 62 12,520 750 5 58 12,363 7 112 18,446 1000 8 72 17,424 9 148 24,460 1500 11 116 24,728 13 223 36,890 2000 15 115 32,970 17 298 49,190 3000 21 235 49,151 26 449 73,785 F.J. Segura-Morales et al. Bioresource Technology 407 (2024) 131147 6 highlights K. cf. nitens as being an exception. Routine examination under an optical microscope revealed the presence of shorter filaments (1–4 cell each) in cultures subjected to high gas flow rates whereas larger filaments (50–100 cells each) were found in static and low gas flow rate cultures (see supplementary material). Notably, filament fragmentation positively affected the growth rate and biomass productivity. This observation is consistent with previous findings for other Klebsormidium species. Mikhailyuk et al. (2014) showed that Klebsormidium species possess a unique cell wall ultrastructure containing a high callose content yet relatively poor in cellulose (Holzinger and Karsten, 2013). Callose provides the walls with considerable flexibility, which may enhance filament fragmentation (Herburger and Holzinger, 2015). According to Stoyneva-G¨ artner et al. (2019), vegetative multiplication through the fragmentation of filaments into smaller and more metabolically active parts, which lack distinguishable polarity, represents the prevailing mode of cell propagation. This rapid filament fragmentation may be a life strategy to better survive in soil crusts; it has been regarded as a biotechnological advance in Klebsormidium strains for mass culture application (Stoyneva-G¨ artner et al., 2019). Therefore, the results obtained in this work from the pneumaticallyagitated flask cultures strongly indicate that K. cf. nitens is robust enough to thrive in culture environments characterized by high fluid-dynamic stress, positioning this microalga as an ideal candidate for mass cultivation in systems where exposure to high hydrodynamic forces is likely, such as in T-PBRs driven by centrifugal pumps (as discussed later in this work). 3.2. Selection of culture medium The culture medium’s composition critically affects microalgal cell growth and the productivity of biomass and products of interest (de Carvalho et al., 2019). For this reason, before studying the potential of K. cf. nitens to grow in the photobioreactors, a brief study was conducted to select the appropriate culture medium for enhancing its cell growth and biomass productivity when scaling up the process, using the culture media described in Section 2.4 and the f/2 medium as the control. As noted in Section 2.4, the culture media f/2 ×6, L1-Se ×6, K-PO 4 ×6 and 3NBBM+V all had the same nitrate concentration. The phosphate concentration was higher in f/2 ×6 and 3NBBM+V. L1-Se ×6 is a modification of the f/2 medium with a broader trace metal composition, including Se, Ni, V, and Cr, elements that are not present in the other media. Fig. 4 shows the variations in the biomass concentration as well as the nitrate and phosphate concentrations throughout the culture period. The cultures were initially started in discontinuous mode. This continued until day 10, after which semicontinuous cultures were initiated, encompassing a total of 6 subcultures up until day 28. The effect of the culture medium formulation on the final biomass concentration during the batch culture period was statistically significant (Fratio =39.6, p<0.05). All the formulations provided favourable conditions for K. cf. nitens growth, as evidenced by the increasing biomass concentration in each culture over time. As can be observed in Fig. 4, a similar growth pattern was seen in all the batch culture phases (until day 10), resulting in final biomass concentrations ranging from a minimum of 210 mg L -1 for the f/2 ×1 medium to a maximum of 500 mg L -1 for the f/2 ×6 medium. Similarly, during the subsequent semicontinuous culture phase (days 10 to 28), there were statistically significant differences between the final biomass concentrations of the different culture media (F-ratio =41.2, p<0.05). While the f/2 ×6 medium, the formulation with the highest macronutrient concentrations, yielded the maximum final biomass concentration and productivity (1160 mg L -1 and 40 mg L -1 day −1 , respectively), the more complex medium (L1-Se ×6), which contained more micronutrients but less phosphates, had a lower biomass concentration and productivity (920 mg L -1 and 30 mg L -1 day −1 , respectively). It is evident that nitrogen limitation only occurred in the culture conducted with the f/2 medium, which contained a lower initial concentration. In contrast, phosphorus limitation was observed in cultures using the f/2, L1-Se ×6 and K-PO 4 ×6 media, which was expected given their lower initial phosphate concentration. In the culture with 3NBBM+V, the low concentrations of essential trace metals, biotin and cobalamin appear to be potential factors contributing to the lower final biomass productivity observed. Interestingly, K. cf. nitens exhibited better growth when there were no Se, Ni, V, and Cr present, exclusively in the L1-Se ×6 medium. Clearly, these additional trace metals were not essential for K. cf nitens. Evidently, different algae have different trace metal requirements. In this regard, F´ abregas et al. (2000) found that Zn, I, B, and V were Fig. 3. Specific growth rate during the exponential growth phase (µ max ) and maximum biomass productivity (P b ) as a function of the gas flow rate. Data points are averages and vertical bars are standard deviations (SD) for samples from duplicate cultures. Values denoted by a different lowercase at each point differed significantly at p<0.05 in the one-way ANOVA. F.J. Segura-Morales et al. Bioresource Technology 407 (2024) 131147 7 detrimental to the growth of Haematococcus pluvialis, whereas Se had no effect on growth at all. Conversely, Chen et al. (2011) found that Dunaliella salina growth was reduced by up to 50 % in the absence of Fe, Mn, or Co. In our study, nutrient deficiencies definitely resulted in decreased cell growth and biomass productivity, while the optimal medium (f/2 ×6) supported maximal K. cf. nitens growth and biomass productivity. Based on the summarized results in Fig. 4, f/2 ×6 was selected as the most suitable growth medium to subsequently culture K. cf. nitens in the T-PBR since it provided both high biomass concentration and productivity while ensuring there was no limitation by nutrients. 3.3. Culture in the photobioreactors T-PBRs have long been recognized as some of the most suitable closed systems for large-scale microalgae cultivation due to their superior surface-to-volume ratio, which ensures effective light absorption Fig. 4. Temporal changes in the cell concentration, and the dissolved nitrate and dissolved phosphate concentrations in the supernatant for the different culture media assayed. Data points are averages and vertical bars are standard deviations (SD) for samples from duplicate cultures. Values denoted by a different lowercase at each point differed significantly at p<0.05 in the one-way ANOVA carried out on the final biomass concentrations during the batch culture phase and the semicontinuous culture phase. F.J. Segura-Morales et al. Bioresource Technology 407 (2024) 131147 8 (Aci´ en-Fern´ andez et al., 2001; Macías-de la Rosa et al., 2023). In this work, the T-PBR described in the Material and Methods section was used to study the scalability of the K. cf. nitens culture. Fig. 5a and b present the culture profile obtained. The F V /F M value did not change significantly over the culture period (p>0.05), with the average value of 0.66 ±0.04 being indicative of healthy cells (Fig. 5a). Regarding the variation in the biomass concentration, there was a statistically significant difference between the means of the three sets (F-ratio =32.1, p<0.05): 485 ±270 mg L -1 , 1845 ±256 mg L -1 and 3385 ±244 mg L -1 , for set 1, set 2 and set 3, respectively. Multiple Range Tests revealed that all means were significantly different from the others (p<0.05). Given the significant differences between the conditions in the system used to maintain the inocula and those in the T-PBR, set 1 was expected to promote K. cf. nitens acclimatisation to the environmental conditions prevailing inside the T-PBR. As shown in Fig. 5a, in this set 1, performed at I 0 =600 µE m −2 s −1 (12:12 h), the biomass increased steadily over time, reaching 1000 mg L -1 at the end of the set (day 14); this value is similar to that achieved in the previous flask culture with the f/2 ×6 medium on day 22. These results indicate that K. cf. nitens successfully adapted to the T-PBR conditions during set 1. After 14 days of cultivation, a slowdown in growth was observed, mainly due to nitrate depletion by day 8 (see Fig. 5b), together with likely limitations in irradiance availability. However, phosphate levels remained in excess at a constant concentration >100 µM after day 9 (see Fig. 5b). On day 15, the culture was changed to semicontinuous mode (set 2), with 3000 mL of culture periodically replaced with nutrient-rich medium, as described in Section 2.5. In addition, the irradiance regime was changed to 600 µE m −2 s −1 (18:6h), which increased irradiance availability to the cells. Set 2 lasted 14 days (until day 29), reaching a biomass concentration of 2800 mg L -1 with concomitant depletion of nitrate and phosphate occurring within 7 days of cultivation (see Fig. 5b). Clearly, the extension of the light period from 12:12 h to 18:6h Fig. 5. Dynamics for the sequential culture of K. cf. nitens in the pilot-scale tubular photobioreactor. Temporal changes in cell concentration and F V /F M (a), dissolved nitrate and phosphate concentrations in the supernatant (b), variation in the concentration of the xanthophyll-cycle carotenoids and the de-epoxidation state expressed as (ant +zea)/(ant +zea +viol) (c), and the fatty acid content of the biomass at the end of each set (d). Data points are averages and vertical bars are standard deviation (SD). In (c) and (d) samples were taken from the final three days of culture for each set, during the culture’s steady state. Values denoted by a different lowercase at each point differed significantly at p<0.05 in the one-way ANOVA. Duplicate samples from each culture time were used. F.J. Segura-Morales et al. Bioresource Technology 407 (2024) 131147 9 in this set resulted in increased nutrient consumption and biomass production, suggesting light-limited growth. Therefore, a subsequent semicontinuous culture was performed under continuous irradiance (600 µE m −2 s −1 ) from day 29 to 43 (set 3), which led to a remarkable increase in the biomass concentration of up to 4700 mg L -1 . In this set, the nitrate was exhausted by day 6 of the culture. The course of the nitrate and phosphate concentrations (Fig. 5b) shows that nitrogen depletion occurred during the early days of each subculture. Interestingly, this depletion was accelerated in subcultures exposed to higher irradiance, suggesting that continuous illumination positively influenced nitrate assimilation. Furthermore, in all three sets, cell concentration continued to increase even after complete nitrate depletion, a phenomenon well documented in microalgae due to luxury nutrient uptake (Lavín and Lourenço, 2005). Regarding phosphate, complete depletion was not observed in any of the subcultures (Fig. 5b). As previously reported (Palabhanvi et al., 2014), this is indicative of conditions that are primarily nitrate-limited. This divergence from the nitrate uptake pattern indicates the absence of luxury uptake with phosphate, highlighting that K. cf. nitens metabolized phosphate, in parallel with its consumption, for the synthesis of cellular constituents (Sciandra, 1991). Available irradiance was another growth-limiting factor. Previous works have revealed that members of Klebsormidiophyceae need irradiance as low as 30 µE m −2 s −1 for optimal growth (Karsten and Rindi, 2010). However, they also exhibit low photoinhibition at high light intensities. This has been interpreted as a powerful photophysiological plasticity (Míguez et al., 2020). Clearly, in this work, the 12:12 h light–dark cycle (set 1) restricted K. cf. nitens growth more than the 18:6h and 24:0h cycles. Such a response is consistent with the documented literature (see Xu et al., 2021) and is of benefit to biomass production in outdoor cultures. As with the cultures grown in the flasks described in Section 3.1, when routinely observed under the microscope, the T-PBR culture showed no evidence that K. cf. nitens growth was negatively affected by the shear stress present in the T-PBR, which operated at 700 L h −1 (corresponding to a fluid circulation velocity of 70 cm s −1 ). This pumping speed led to an average shear stress of 6600 mPa, derived from local values of 1786 Pa for the pump, 2800 mPa for the degasser, and 600 mPa for the light collector. As the cells passed through the pump, they were subjected to shear stress that was three to four orders of magnitude greater than the levels they encountered in either the degasser or the loop. However, the residence time in the pump was only 150 ms compared to 3850 ms in the degasser and 43 s in the loop during the cycle time (47 s). The high shear stress the microalga was subjected to as it passed through the pump impacted the intercellular junction more than other cell structures. This caused filament fragmentation, thus limiting the length of the filaments, which in most cases consisted of 1 to 4 cells. The same phenomenon was found in the cultures subjected to the highest gas flow rates described in Section 3.1 although some larger filaments were observed there. Such filament breakage clearly favours the vegetative multiplication of K. cf. nitens. Regarding the O 2 concentration in the T-PBR, this was monitored online as an indicator of photosynthetic activity. Nonetheless, due to the high fluid circulation velocity (70 cm s −1 ) and the limited length of the light collector (30 m), the photosynthetic dissolved O 2 that accumulated never reached levels that could cause growth inhibition or photooxidation −the O 2 concentrations measured at the inlet of the light collector tube were always close to the air saturation value and those measured at the outlet were never greater than 120 % of air saturation, far below the suggested healthy upper limit of 250 % air saturation (Aci´ en-Fern´ andez et al., 2001). During the dark phase, when there was no illumination, the O 2 concentration never dropped below 50 % of air saturation. Along with T-PBRs, bubble columns are amongst the most preferred photobioreactor configurations because they are comparatively easy to design and operate, and have no moving parts. They have been extensively used to grow a range of microalgae (see L´ opez-Rosales et al., 2019) since the environment only exposes the microalgae to mild shear stress (Manjrekar et al., 2017). In this work, the BC-PBR described in the Material and Methods section was also used as an alternative for scalingup the cultivation of K. nitens. The air flow rate in the BC-PBR was set at 13 L min −1 , matching the rate that maximized the µ max and P b in the preliminary culture flask experiments (1.3 vvmin). Despite the apparent health of the cells, as indicated by the F V /F M measurements taken throughout the culture period (0.64 ±0.02), the culture was sustained only during the initial two sets (30 days of culture) due to exceptionally erratic biomass concentration measurements (data not shown). This instability stemmed from the buoyancy properties of K. cf. nitens and its morphological response to the pattern of hydrodynamic forces existing in the BC-PBR. This resulted in a highly heterogeneous distribution of cells within the culture, with much of the suspended biomass accumulating in the region just below the culture surface, forming dense clumps and mats, and with significant cell adhesion to various surfaces inside the column (see supplementary material). The biomass concentrations in the culture at the end of sets 1 and 2 were determined to be roughly the same −at around 400 mg L -1 −well below those achieved in the T-PBR (4700 mg L -1 ). In bubble columns with a high AR, such as the BC-PBR used in this study (AR=21), the flow pattern tends to be highly stratified, with the maximum axis fluid velocity being just above the sparger; this decreases as the distance from the sparger increases (L´ opez-Rosales et al., 2019). Such a slow velocity near the top creates a region of reduced shear stress, which is favourable for the accumulation of filamentous microalgae that aggregate and form mats or biofilms. Additionally, bubble-column photobioreactors rely on gas bubbles to provide aeration and mixing. Taller columns allow more time for the microalgae to attach and accumulate on bubble surfaces and move up with the rising bubbles, contributing to the non-uniform distribution of cells and to the formation of clumps and mats towards the top of the column (Ding et al., 2016). Furthermore, the presence of gas bubbles or air pockets within the filaments can further enhance their buoyancy. Conversely, in photobioreactors with a low AR, such as the spherical culture flasks used in Sections 3.1 and 3.2 (with an AR of around 1), the flow is more turbulent, and mixing is more efficient due to the shorter distances the bubbles travel. This leads to a more homogeneous distribution of microalgae throughout the vessel, with less opportunity for filamentous microalgae to accumulate at the top. Understanding the buoyancy of filamentous microalgae is clearly important in designing, optimizing, and selecting photobioreactors for their culture. Therefore, the challenge of achieving a homogeneous K. cf. nitens cell culture in tall bubble columns raises doubts regarding the suitability of this Klebsormidium strain for large-scale cultivation. This finding is contrary to that obtained by other authors, who reported that the Klebsormidium sp. LGX80 strain grew much better in a BC-PBR (AR=13) than in a T-PBR driven by an airlift-pump (Xu et al., 2021). The authors observed in the T-PBR that during the rapid growth phase of Klebsormidium sp. LGX80, the filamentous cells tended to entwine tightly, forming clots or mats that eventually disintegrated into shorter filaments. They suggested that this was most likely due to the shear stress from the air-pump system or air bubbles, along with the thin cell walls. However, these observations were limited to a very short period within the 32-day culture, and the authors did not provide further information on potential biofouling formation. Therefore, it is difficult to make conclusive comparisons with this study. Despite this, it is apparent that considerable variability exists between different Klebsormidium species when cultivated in photobioreactors that have typical configurations. Overall, the design of the centrifugal pump-driven T-PBR used in this study offers clear advantages for growing the filamentous microalga K. cf. nitens. The cells are subjected to short but intense stress as they pass through the pump, followed by prolonged low-stress conditions within F.J. Segura-Morales et al.