Effects of cryopreservation on viability and functional stability of an industrially relevant alga

As algal biotechnology develops, there is an increasing requirement to conserve cultures without the cost, time and genetic stability implications of conventional serial transfers, including issues regarding potential loss by failure to regrow, contamination on transfer, mix up and/or errors in the documentation on transfer. Furthermore, it is crucial to ensure both viability and functionality are retained by stored stock-cultures. Low temperature storage, ranging from the use of domestic freezers to storage under liquid nitrogen, is widely being used, but the implication to stability and function rarely investigated. We report for the first time, retention of functionality in the maintenance of master stock-cultures of an industrially relevant, lipid-producing alga, under a variety of cryopreservation regimes. Storage in domestic (−15 °C), or conventional −80 °C freezers was suboptimal, with a rapid reduction in viability observed for samples at −15 °C and a >50% loss of viability, within one month, for samples stored at −80 °C. No reduction in viability occurred at −196 °C. Post-thaw culture functional performance was also influenced by the cryopreservation approach employed. Only samples held at −196 °C responded to nitrogen limitation in terms of growth characteristics and biochemical profiles (lipid production and chlorophyll a) comparable to the untreated control, cultured prior to cryopreservation. These results have important implications in microbial biotechnology, especially for those responsible for the conservation of genetic resources.


Results
Effects of cooling regime and storage on C. vulgaris viability. A graphical description of the experimental workflow can be found in Fig. 1, including details of the cryopreservation treatments and the physiological experiment carried out to test functionality. Simple re-growth assay of the cultures was unaffected by duration of storage up to 4 months at −80 °C and −196 °C (Table 1). However, re-growth of cultures was impacted at −15 °C after only 1 month of storage. This indicated that −15 °C approach was fundamentally unreliable. In terms of absolute viability levels, cells held at −196 °C maintained near 100% viability for 4 months. In contrast, significant reductions (from 98% to 28%) were seen after 1 and 4 months at −80 °C (suggesting a half-life of 1-2 months) and viability was lost after 1 month at −15 °C as judged by the pour-plate assay. Therefore, a closer analysis at the cellular level suggested that there were long-term limitations in the use of the 'lab-freezer' approaches, favouring the cryopreservation approach.
Post-storage performance, functional phenotype of C. vulgaris. The  showed that following the removal of nitrogen from the media at the beginning of Phase 2, no inorganic nitrogen was detected in the media (up to 32 h for the case of −15 °C and −80 °C) and that all the cryopreserved treatments continued to be nitrogen starved during the duration of the physiological experiment ( Fig. 2A). −15 °C and Figure 1. Experimental workflow adopted. The experiment consisted of testing three cryopreservation strategies. This was preceded by a physiological experiment to assess functionality prior to cryopreservation and consisted of growth in two phases (A), and subsequent preparation of samples for cryopreservation (B). The cryopreservation step (C) was then followed by thawing and culture recovery (D) and viability assessment and culture acclimatisation (E), before repeating the physiological experiment on the cultures post-cryopreservation in two phases (F) to test for functionality. −80 °C cryopreserved strains showed a slight increase in the DIN from 32 h onwards, perhaps, as a result of remineralisation of organic nitrogen released into the media by these cultures. The DIP was consumed in the first 4 hours of nitrogen starvation in all treatments (Fig. 2B). After recovery from the cryopreservation treatments, both −15 °C and −80 °C stored cultures were slower in picking up growth under the replete nutrient conditions of Phase 1 (0-24 h) when compared to the −196 °C treatment (p-value < 0.001) (Fig. 3A). This was noted despite equivalent prior acclimatisation (Fig. 1). As expected, the initial growth rates (0-4 h) for Phase 2 of all the cultures under nitrogen starvation (the Control N free and the three cryopreserved cultures) were slower than the initial growth rate observed for the Control N replete, suggesting the requirement for a certain time to acclimatise the    Fig. S1A). In our study, the carbon to nitrogen (C/N) ratio was measured as a proxy to indicate the degree of nitrogen limitation within the algal cells. The C/N molar ratio rapidly increased to values higher than 10 at 24 h in the −196 °C cryopreserved populations and in the Control N free treatment. For the −15 °C and −80 °C stored populations, the increase was slower reaching a C/N of 10 at 48 h (Fig. 4A). The Control N replete C. vulgaris population did not show signs of intracellular nitrogen limitation until the end of the experimentation when the nitrogen in the medium was exhausted. The intracellular particulate organic carbon (POC) increased by 20-60% of the initial value over 72 h in all the treatments, except the −15 °C stored culture, where the increase was minimal at around 10% (Fig. 4B). A concomitant decrease in intracellular particulate organic nitrogen (PON) was noticed in all the treatments, with a dramatic drop to 60% in the first 24 h and a further drop to 40% of the initial value in 72 h in the −196 °C cryopreserved and 50% in Control N free populations, contrasting with the more gentle drop in the −15 °C and −80 °C stored cultures that reached values that were 60% of the initial values at 72 h, more similar to that noticed with the Control N replete condition (Fig. 4C). The cultures stored at −15 °C and −80 °C thus seemed to have a better ability to retain the intracellular nitrogen than the −196 °C cryopreserved and Control N free populations, showing both higher biomass productivity (indicated by having ODs 50% higher from 48 h onwards - Supplementary Fig. S1A) and higher PON levels which were between >50% higher from 24 h until the end of the experiment. No noticeable increase of DIN in −196 °C stored and Control N free . The specific growth rates were measured as μ = ln(X t /X 0 )/(t t − t 0 ), X t and X 0 being the biomass concentrations (as OD 595nm ) at time points t hours (t t ) and 0 hours (t 0 ), respectively. Mean (±standard error) values of three biological replicates are plotted. Analysis of the variance followed by the Bonferroni post hoc test (p-value < 0.05) was carried out to estimate the significance of the differences between the treatments (i.e., (a) significant difference when compared with Control N replete, (b) significant difference when compared with Control N free, (c) significant difference when compared with −15 °C treated populations, (d) significant difference when compared with −80 °C treated populations and (e) significant difference when compared with −196 °C treated populations). In (B) the ANOVA between treatments was carried out for the maximum growth rates and between the "Initial" and the "Maximum" growth rates within the same treatment, (*) meaning significant difference between the two types of rate for each treatment.  Fig. 2A). The −15 °C and −80 °C stored cultures also presented differences in their biochemical composition with phenotypes that were between an intracellular nitrogen limited and replete scenario, at least during the first 48 h of the nitrogen starvation. For instance, the −196 °C stored culture was observed to have a decrease in the chlorophyll a and total protein similar to that in the Control N free culture (Fig. 5A,B), which contrasted with the relatively unchanged concentrations in the Control N replete condition. Conversely, the −15 °C and −80 °C stored cultures had slower decreasing trends. Similar behaviour was observed with respect to lipids (Fig. 5C). Both Control N free and −196 °C stored cultures presented a steady increase in the lipid content from the start of the nitrogen starvation, although differences in the degree of accumulation were observed (5-fold and 3.5-fold, respectively). Finally, the −15 °C and −80 °C stored cultures clearly had their ability to accumulate lipids compromised by the cryopreservation storage temperatures employed, as their trends were similar to the changes observed for the Control N replete treatment.

Discussion
Whilst the use of suboptimal environmental conditions including incubation temperature and illumination (for low growth) is the standard approach in maintaining live algal cultures 9 , it is considered inadequate in the context of biotechnologically exploited materials in terms of both strain security and ensuring culture "performance" 20,30 . Where effective methodologies have been developed, low temperature storage, specifically ultra-low temperature cryopreservation, is accepted as the optimal approach 8,20,30 . The tolerance of biological materials to cryopreservation depends upon the ability of the material/organism being cooled to overcome or avoid cryo-injury induced by physical and chemical changes associated with freezing of the surrounding medium and the material/organisms themselves. Mazur 31 proposed that two factors were involved namely: ice formation and colligative damage due to excessive concentrations of solutes as ice is formed. There is a considerable body of work exploring both the physics of cryopreservation and its effects on biological systems, for comprehensive reviews see Taylor (1987) 32 and Muldrew et al. 33 , respectively. From a practical perspective, for many biological materials (from viruses to mammalian embryos) standard methodologies applicable to a range of organisms/materials have been developed 34,35 . However, for algae, as for other groups of organisms, it is clear that there is no universally applicable method, although "standard" protocols, applicable for a variety of cyanobacteria 19 , microalgae 36 and macroalgae 37 , have been published. Irrespective of the methodological approach, stability of the stored samples is of paramount importance. In theory, where storage temperatures are sufficiently low to prevent molecular motion, cryogenically stored samples should be effectively inert. Thus, the −196 °C preserved cells are in an "arrested" state and long-term storage should result in little or no loss of viability. The key to this stability is the maintenance of samples in a glassy (vitrified) state, where neither solute/salt nor ice-induced injuries can occur. To achieve this, cryopreserved samples need to be stored at temperatures less than the critical glass transition temperature (T g ), which for pure water has been reported as being between −130 °C and −132 °C 38 . Wowk (2010) 39 reported that T g is relevant to the stability of cryopreserved samples in the totally vitrified state and in the frozen state, where cells survive in glasses between ice crystals. In the majority of samples that are thought of as frozen (i.e. with a large content of crystalline ice), rather than vitrified (i.e. water and other components form a glass) there are assumed to be regions within the sample where a 'glassy' matrix has been formed 40 .
In common with other biotechnological sectors, few algal biotechnologists have a significant level of cryobiological experience and are unaware of the potential implications of employing different storage temperatures. Furthermore, although sample stability is critical, an equally important consideration is that methods should be practical. This had led to a number of workers exploring the potential of employing domestic freezers 41,42 , as these are readily available in virtually all labs and production facilities. Furthermore, −80 °C storage is the standard approach for the conservation of bacterial cultures 27 and has been successfully employed for a variety of eukaryotic organisms 8,34 . Irrespective of the protocols employed, the objectives of any cryopreservation procedure are twofold: firstly, to have viable material capable of regenerating a new culture, plant or animal and secondly, the revived material should regenerate as phenotypically and functionally "normal" i.e. stored revived materials must be "fit for purpose". In this study, we systematically investigated the effects of cryopreservation and storage of a lipid producing alga C. vulgaris CCAP 211/21A. This alga was selected because it is the model strain on which a substantial body of research has been performed in our labs 29,43,44 and we were confident of the stability of the alga under its standard maintenance regime. In addition, its relative C. vulgaris CCAP 211/11B, which is employed worldwide in ecotoxicology 45 , has been demonstrated to retain its genotypic stability for >40 years on being serially transferred under different cultivation regimes and under liquid nitrogen storage 10 . The storage temperatures were chosen because they reflect equipment availability and we selected a −15 °C, rather than a −20 °C, domestic freezer as we wanted to explore the effect of the highest sub-zero temperature that would ensure that the material, or its external medium, were frozen. It is worth noting that samples in marine-based algal media may supercool to <−12 °C before extracellular ice nucleation occurs 11 . The period of 4 months was chosen as this is the shortest duration of storage that is less labour intensive in comparison to routine serial transfer for this alga.
On storage at −15 °C, we observed an unacceptably rapid reduction in viability of C. vulgaris CCAP 211/21A to <1% and failure to revive some samples within one month's storage. This is similar to an earlier report 46 , where green spores of the ferns Osmuncla regalis and Equisetum ramosissimum lost viability quickly, with unacceptably high levels of variability on storage at −25 °C over a 24-month storage period. In this study, storage at −80 °C was suboptimal with a 70% reduction in viability levels, although all replicates tested up to 4 months storage revived and regenerated actively growing cultures. It may be assumed that the reduction in viability was due to damage to algal cells associated with ice crystal development and other physico-chemical changes in the ice matrix at this relatively high sub-zero storage temperature. For some eukaryotic organisms, −80 °C storage has proven to be satisfactory such as, for example, dried spores of O. regalis and E. ramosissimum which maintained high viability, rapid germination and normal growth over 24 months storage 46 . Furthermore, there are published examples of successful short-term (weeks-months) storage of eukaryotic algae at −80 °C 47,48 . As might be anticipated, based on cryobiological theory, materials held at −196 °C showed no reduction in viability over the four months storage. Current cryobiological understanding is that cells capable of functional recovery exist in the vitreous state at ultra-low temperatures, even though they may be surrounded by a 'wall' of ice 40 . This explains why storage below the critical T g (~ −130 °C) is crucial to ensure viability for both frozen and totally vitrified samples. In this study, the samples that were plunged into liquid nitrogen would have contained vitrified cells within a frozen matrix. Samples subsequently stored at −80 °C will have devitrified when transferred from liquid nitrogen to the −80 °C freezer, this in itself did not result in any reduction in viability levels. However, on prolonged storage intracellular ice formation and salt induced injuries occurred in many algal cells, with lower recovery levels observed on thawing of the cryopreserved samples.
When conducting cryopreservation of living materials, in addition to testing for the viability of the storage cultures, it is equally important to determine that the functionality of the revived material is not lost and therefore is "still fit for purpose". This is a fundamental requirement in biotechnology where maintaining a consistent phenotype and functionality is critical to keep a consistent commercial exploitation. However, for most practitioners, the approach employed is typically empirical (often copying a method used for some other organism), and there is no attempt to provide an adequately qualified reference stock of material to serve the increasingly sophisticated analytical systems for characterizing the sample 49 . Thus, there is often a lack of assurance that culture recovered in the future will have retained the properties of the original cells. The use of suboptimal storage regimes may further compound this issue, but to date, we are unaware of any reports of this being systematically investigated for algae. Ryan et al. 50 reported that storage at −196 °C rather than −20 °C was preferable for the conservation of insecticidal secondary metabolite production by the fungi Metarhizium anisopliae and Fusarium oxysporum. More recently, evidence of molecular changes was reported to 18 genes associated with stress physiology, on storage of Human Peripheral Blood Mononuclear Cells (PBMCs) at −80 °C, but no changes were observed in samples stored <−150 °C 51 . Furthermore, as noted for C. vulgaris, PBMC recovery and viability were better when the cells were stored ≤−150 °C as compared to −80 °C 51 .
In our study, we assessed the effect of the different cryopreservation storage temperatures on the physiology of C. vulgaris, with a specific focus on the changes observed in its response to nitrogen starvation. This environmental condition is widely used in the algae biotechnology field to stimulate lipid production required for the generation of biofuels 29,52 . To achieve that goal, a physiological experiment involving cultivation under conditions of nitrogen starvation in the medium was conducted on algal cultures; before (Control N free) and after the cryopreservation and storage at the different temperatures (−15 °C, −80 °C and −196 °C), and any changes in the biochemical composition observed were compared to the growth of the same alga in the presence of nitrogen (Control N replete). The physiological responses to the nitrogen starvation were very similar in both algal cultures before (Control N free) and after the cryopreservation with storage at −196 °C. For instance, compared to the Control N replete culture, maximal growth rates were significantly lower in both Control N free and −196 °C derived cultures, with their growth arrested at 24 h, and the molar C/N ratio, a proxy of the intracellular nitrogen limitation, increased to levels up to 25 by the end of the physiological experiment. In addition, the decrease of chlorophyll a and increase of lipid content trends were very similar for both Control N free and −196 °C stored cultures and very different to that observed for the Control N replete cultures. These results coincide with what has been described for this and other species of algae subjected to the stress of nitrogen starvation 53,54 , and further endorse cryopreservation and storage at −196 °C for the long-term maintenance of C. vulgaris without compromising its functionality.
Contrasting to the results observed for samples stored at −196 °C, both −15 °C and −80 °C derived cultures showed different physiological responses to nitrogen starvation when compared to the Control N free treatment, having phenotypes that were between an intracellular nitrogen limited and replete scenario. For instance, they seemed to be more productive than the −196 °C and Control N free cultures, their growth was not arrested until towards the end of the physiological experiment, and their C/N ratios increased very slowly over time. In addition, although they showed a decreasing trend in their chlorophyll a and total protein content, the lipid content was very similar to the Control N replete culture, indicating that lipid accumulation was not stimulated, at least for the duration of the experiment. Finally, they seemed to release nitrogen to the environment, an ability that has been reported for different microalgae species both in the nature and in culture. All these differences together with the presence of a lag in the growth during the Phase 1 suggest that a sub-population of C. vulgaris had been selected through the cryopreservation/storage procedures at −15 °C and −80 °C and that cells with higher biomass productivity could be emerging from a sub-population of viable cells. This sub-population presented a new physiological phenotype, no longer directly comparable to the original culture, which was able to maintain growth in the absence of nitrogen for a longer period of time. The fact that this new phenotype presented higher PON levels than the −196 °C and Control N free cultures also pointed towards its physiological differentiation, as it contrasts with the loss of intracellular nitrogen due to organic matter secretion that has been previously described for chlorophyceans 55 .
In conclusion, this paper has provided important evidence that storage at high sub-zero temperatures cannot guarantee the stability of either viability or function for the alga studied. Storage at −15 °C and −80 °C seemed to select for a sub-population of C. vulgaris with very different physiological properties (able to keep growing for longer periods in the absence of nitrogen) than the original isolate. This change in functionality has profound biotechnological implications for the conservation of algae and other biological resources for biotechnology. It is advised that for commercially viable isolates, in particular, that conservation/cryopreservation methodologies should be optimised and functionality validated before the approach is used as standard practice.

Methods
Culture and routine maintenance regime. C. vulgaris CCAP 211/21A obtained from the Culture Collection of Algae and Protozoa was routinely maintained by the aseptic transfer of a 5% (v/v) inoculum into a 50 mL tissue culture flask, with a vent closure permitting gas exchange, containing 30 mL f/2 medium prepared according to Guillard (1975) 56  Experimental design. To study the effect of different cryopreservation protocols on the physiological responses of C. vulgaris to nitrogen starvation, the same experiment was performed on four occasions. This involved studying triplicate samples prior to cryopreservation, this acted as a control, or initial characterization of the physiological status of the strain; and then on populations/cultures recovered from three cryopreservation treatments (see below). The physiological experiment was carried out in two phases (Fig. 1) sparger, pore size 3, at a rate of 0.01 L m −1 . The cultures were incubated at room temperature under continuous light at 200 µmol photons m −2 s −1 provided by cool white LED stripes. Initial inoculum density was at an optical density at 595 nm (OD 595 ) of 0.15 (equivalent to ca. 10 6 cells/mL) and growth was monitored daily by measuring OD 595 . When OD 595 was in the range 0.5-0.6, the 3 × 1 L cultures were mixed and concentrated by centrifugation at 3500 g for 5 min in an Eppendorf 5810 centrifuge before being re-suspended to an equivalent OD 595 of 0.6 (±0.1), as triplicates, in 1 L of f/2 medium free of nitrogen, which marked the start of the experimental Phase 2. In Phase 2, the same culture conditions were maintained with the exception of increasing the light to 250 µmol photons m −2 s −1 . A similar physiological experiment using a normal f/2 medium in the second phase was performed prior to cryopreservation, to assess the growth under replete nutrient concentrations. Samples for determining cellular biochemical and elemental composition were taken at different time points. In addition, culture biomass was determined by measuring the OD 595 .
Cryopreservation experiments. After the first (control) physiological experiment, two 15 mL samples of C. vulgaris were sent to the Scottish Association of Marine Sciences to conduct the cryopreservation experiments. On receipt, these were incubated for 24 h under the standard maintenance conditions outlined above. They were then combined and then three aliquots were subjected to three cooling and storage regimes.  Thawing and culture recovery. After storage, the vials were rapidly transferred to a 37 °C water bath and placed in floating vial holders. As soon as all ice had melted, 1 mL from each vial was aseptically transferred into a 50 mL tissue culture flask containing 25 mL of f/2 medium. The cultures were allowed to recover in complete darkness for 24 h and then incubated at 20 °C, under a 12 h:12 h light-dark photoperiod at 20 μmol photons m 2 s −1 .
Viability assessment. Viability levels were assessed on the basis of capacity to grow in liquid culture and to form discrete colonies in agar, according to the method of Day and Deville (1995) 57 . In brief, aliquots (1 mL) of logarithmic dilutions in sterile medium diluted to result in 50-500 cells mL −1 were aseptically transferred to pre-labelled Petri dishes (50 mm diameter) to which approximately 2.5 mL of medium (f/2) containing agar (1% w/v) was dispensed. After gelation of the agar, the plates were sealed with Parafilm and incubated, inverted, under standard algal culture conditions (see above) for 4 weeks. This procedure was repeated for thawed samples after 24 h, 1 month and 4 months storage. After 4 weeks incubation the agar plates were examined under a dissecting microscope (x 20 magnification), the number of colonies counted and viability expressed as a percentage of the original cell density of the control sample. Cultures recovered after 4 months storage were incubated under the standard maintenance regime until a dense culture was obtained. These were then returned to the University of Sheffield for further study. On receipt, they were aseptically transferred to 1 M Tris buffered f/2 medium where they were acclimated to the laboratory conditions by incubation at room temperature in 250 mL flasks aerated with 0.22 µm filtered moisture air at 200 µmol photons m −2 s −1 irradiance. The recovered samples were maintained under these conditions in an exponential growth phase for 16 (−15 °C and −80 °C treated) and 22 (−196 °C) passages (replacing the medium every 4-6 days) before repeating the physiological experiment described above.
Determination of DIN and DIP in the medium. After centrifuging the samples for the determination of the biochemical composition (see details below), supernatants were frozen and stored at −20 °C for further analysis of the medium nutrient composition. DIN in the media was determined by measuring the absorbance at 220 nm as described in Collos et al. 58 . Soluble Reactive Phosphate in the media (DIP) was determined following the method of Strickland and Parsons's method (1968) 59 .
Determination of the biochemical and elemental composition. All chemicals and analytical reagents were HPLC grade (Sigma-Aldrich, U.K.) unless stated otherwise. Samples (5 mL) from each bottle were centrifuged at 3500 g for 5 min. Pellets were washed with phosphate buffer 0.01 M before being transferred to 2 mL Eppendorf tubes and centrifuged at 7000 g for 2 min. After removing the supernatant, pellets were stored at −80 °C before being freeze-dried. Pigment (Chlorophyll a) and total protein concentrations were determined in Scientific RepoRts | (2019) 9:2093 | https://doi.org/10.1038/s41598-019-38588-6 the 5 mL-equivalent pellets of dry biomass using a modification of the combined assay described in Chen et al. 60 . For determining the Particulate Organic Carbon (POC) and Organic Nitrogen (PON) inside the cells, 15 mL samples were centrifuged and pellets freeze-dried as described above, and known amounts of dry biomass analysed in an Elemental Analyzer (Flash 2000) using L-Isoleucine as a standard.
Determination of Total lipids. All chemicals and analytical reagents were of HPLC grade (Sigma-Aldrich, Dorset, U.K.) unless stated otherwise. A 5 mL algal culture was harvested by centrifugation at 19,000 g for 3 min to which 300 μL of toluene and 300 µL of 0.5 M sodium methoxide were added, followed by incubation at 80 °C for 20 min. After cooling to room temperature, 300 µL of 10% boron tri-fluoride in methanol was added and the mixture incubated at 80 °C for 20 min. After cooling to room temperature, 300 µL water and 600 µL of hexane were added. The mixture was vortexed for 1 min and centrifuged at 18,000 g at 4 °C for 10 min. The organic phase was recovered, measured and evaporated to dryness under inert nitrogen gas. The dried fatty acid methyl esters (FAMEs) were reconstituted in 80 μL hexane prior to identification and quantification as total lipids, as described elsewhere 61 .
Statistical analysis. The statistical differences between the treatments for the physiological parameters determined were analysed by performing an analysis of the variance (ANOVA) followed by the Bonferroni t-test using the SigmaPlot software version 13.

Data Availability
All data generated or analysed during this study are included in this published article (and its Supplementary  Information files).