Isolation, fatty acid profiles and cryopreservation of marine thraustochytrids from mangrove habitats in Thailand

Sample collection and isolation of microorganisms

Decaying mangrove leaves, seagrasses and seawater samples were collected from 22 mangrove habitats located in 12 provinces from central, eastern and southern Thailand (Figure 1). The leaf samples were washed with sterile natural seawater (NSW) containing penicillin G and streptomycin sulfate (500 mg l⁻¹ each, Bio Basic Inc., Canada). These substrata were cut into circular discs (0.5 cm in diameter) and plated directly on the isolation agar medium, Glucose Peptone Yeast Extract (GPY) containing 1 g peptone (Difco Laboratories, Detroit, USA), 2 g yeast extract (Difco), 4 g glucose (Difco) and antibiotics (Yang et al. 2010), prepared in 1 l of half-strength NSW (1:1 NSW/distilled water). To obtain thraustochytrids from seawater, pine pollen was used as a bait and added to the seawater samples (Porter 1990). After 72 h incubation, the pine pollen grains were transferred to new GPY agar plates and subsequently incubated at 25°C. The thraustochytrid-like colonies were aseptically transferred to new agar plates and repeatedly sub-cultured on freshly prepared GPY medium until axenic cultures were obtained. The isolated thraustochytrids were characterized based on morphological features and molecular identification. All axenic cultures were kept in the BIOTEC Culture Collection (BCC), Thailand. Pure cultures were grown in Glucose Yeast Extract (GY) medium containing (per liter of half-strength NSW) 10 g yeast extract and 30 g glucose, and shaken at 200 rpm at 25°C for 7 days. The cultured cells were harvested by centrifugation at 10,640 g, washed twice with sterilized 0.85% NaCl (Carlo Erba, Lombardy, Italy), and lyophilized to determine cell dry weight.

Figure 1:

Map of collection sites in 12 provinces of Thailand.

Morphological observation

Colonies of 26 thraustochytrid strains were sub-cultured using a sterilized inoculating loop, streaked across seawater GPY agar and incubated at 25°C for 3–7 days. Colonies appearing on the plates were transferred to 3 ml seawater GPY broth in six-well plates and incubated at 25°C. Aliquots of the cell suspension (100 μl) were transferred into a slide chamber for microscopic observation. Morphological characteristics of the thraustochytrid strains during cell cycle development, such as presence of ectoplasmic net, formation of zoospores, amoeboid cells, vegetative cell, zoospores and zoosporangia were observed (Yokoyama et al. 2007).

Genomic DNA extraction, PCR amplification and DNA sequencing

Genomic DNA of selected strains was extracted by using DNeasy Plant Mini kit (Qiagen, USA). Partial nuclear small subunit (SSU) rDNA was amplified using primer sets: NS1/NS4 and Thr404f/Thr1017r (White et al. 1990, Honda et al. 1999, Harel et al. 2008). All the PCR primer pair amplifications were carried out using BIORAD T100 Thermal Cycler (USA). Amplification using the primer pair NS1/NS4 was performed as follows: initial denaturation (2 min at 94°C), followed by 35 cycles of denaturation (1 min at 94°C), annealing (1 min at 55°C), extension (2 min at 72°C) and a final extension step (10 min at 72°C; White et al. 1990). Amplification using the primer pair Thr404f/Thr1017r was performed as follows: initial denaturation (4 min at 95°C), followed by 30 cycles of denaturation (30 s at 94°C), annealing (40 s at 50°C), extension (90 s at 72°C) and a final extension step (20 min at 72°C; Harel et al. 2008). The PCR products were then directly sequenced by Macrogen Inc. (Korea).

Sequence alignment and phylogenetic analyses

The nuclear SSU rDNA sequences of the thraustochytrids obtained in this study (Table 1) were aligned with nucleotide sequences retrieved from GenBank, with Bacillaria paxillifera (O. F. Müll.) Hendey M87325 chosen as the outgroup, using the multiple alignment program CLUSTAL W 1.6 (Thompson et al. 1994). The alignment was refined visually in BioEdit 6.0.7 (Hall 2004) and incorporated into PAUP 4.0b10 (Swofford 2003) for phylogenetic analyses through a maximum parsimony (MP) criterion. All characters were equally weighted, followed by heuristic searches with a stepwise starting tree, a random stepwise addition of 10 replicates and tree-bisection-reconnection (TBR) branch-swapping algorithm. The most parsimonious trees (MPT) were tested for the best topology with the Kishino-Hasegawa (K-H) maximum likelihood test (Kishino and Hasegawa 1989) to find the most likely tree for the dataset. Tree length, consistency indices (CI), retention indices (RI) and rescaled consistency indices (RC) were calculated for each tree generated. Finally, 1000 replicates of MP bootstrapping analysis (Felsenstein 1985) were performed through full heuristic searches, stepwise addition of sequence, 10 replicates of random addition of taxa and TBR branch-swapping algorithm.

The Bayesian inference was calculated with Mr. Bayes 3.0b4 with the general time reversible (GTR) model of DNA substitution and a gamma distribution rate variation across sites (Huelsenbeck and Ronquist 2001). Four Markov chains were run from random starting trees for 3,000,000 generations and sampled every 100 generations. The first 300,000 generations were discarded as burn-in of the analysis. A majority rule consensus tree of all remaining trees, as well as the posterior probabilities (PP), was calculated. The maximum-likelihood (ML) analyses were conducted in the CIPRES web portal (Miller et al. 2010) using RAxML 8.2.4 (Stamatakis 2014) with the Broyden-Fletcher-Goldfarb-Shanno (BFGS) method to optimize GTR rate parameters. The maximum-likelihood bootstraps were estimated with the substitution matrix as GTR model and a discrete gamma distribution.

Fatty acid extraction and analysis

Strains cultivated in GY medium at 25°C and 200 rpm for 7 days were harvested by centrifugation and lyophilization. Freeze-dried cells were then directly trans-esterified with 4% sulphuric acid in methanol and then heated in a 90°C water bath for 1 h. Nonadecanoic acid (C19:0) was used as an internal standard (Sigma-Aldrich, St. Louis, MO, USA). The esterified samples were applied to a gas chromatograph (GC17A, Shimadzu), equipped with a 30 m×0.25 mm Omegawax™ 250 fused silica capillary column (Supelco, USA), an automatic sampler and flame ionization detector (FID). The injector and detector temperature were kept at 250°C and 260°C, respectively. Helium was used as a carrier gas at a linear velocity of 30 cm s⁻¹. The initial column temperature of 200°C was held for 10 min and then increased at 20°C min⁻¹ to 230°C where it was held for 17 min. Peaks were identified based on the retention times relative to fatty acid methyl ester standards (Supelco 18919-1 AMP; all from Sigma-Aldrich, USA).

Cryopreservation of selected thraustochytrids

Four marine thraustochytrids (JS510, JS702, JS974 and JS1085) were selected, based on differences in morphological features and phylogenetic position, for cryopreservation at −80°C and vapor-phase liquid nitrogen. The thraustochytrid strains were grown in 50 ml GPY medium at 25°C with shaking at 80 rpm (Cox et al. 2009). Cells were harvested during the log phase growth period (at day 3–5, depending on the species) by centrifugation (10,640 g) for 5 min. The resulting pellet was re-suspended in an equal volume of fresh GPY medium, and the total number of cells was enumerated through serial dilution and a total plate count on GY agar plates. The number of colonies (colony forming unit, CFU) formed per milliliter (CFU ml⁻¹) was calculated for each thraustochytrid strain before freezing.

To determine the most suitable CPA for cryopreservation of selected thraustochytrids in −80°C, an aliquot of the concentrated cells was added to cryotubes containing five different CPA treatments:

1. 5% DMSO (D, Bio Basic Inc., Canada),

2. 10% glycerol (G, Bio Basic Inc., Canada),

3. 17% skim milk (Difco) combined with 20% glycerol (SG),

4. 30% horse serum (GIBCO, USA) combined with 10% DMSO (HD)

5. 5% trehalose (H+B Lifescience, Japan) combined with 10% glycerol (TG).

Cryotubes were placed into a controlled-rate freezing container (Mr. Frosty, Nalgene Cat. No. 5100) and frozen at 1°C min⁻¹ before storage in a −80°C freezer for 6 months. To test the survival of thraustochytrids, the cryotubes retrieved from the −80°C freezer were directly thawed by immersion in a 30°C water bath for 3 min (Nishii and Nakagiri 1991). One micro liter of the contents was serially diluted, and 0.01 ml was then directly spread onto GPY agar plates. The colonies formed on the plate were counted to calculate the proportion of viable cells. The survival rates were calculated relative to the cell density before freezing.

The survival rates of thraustochytrids preserved in liquid nitrogen were subsequently determined by placing the cryotubes containing the concentrated cells in the best CPA obtained from the previous experiment in a vapor-phase of nitrogen for 12 months. The survival of the thraustochytrids was tested in the same way. Statistical significance was established with one-way ANOVA using SPSS 11.5 software for Windows.

Isolation, morphological observation and molecular phylogeny

More than 300 strains of thraustochytrids were isolated in this study. The majority of the isolated strains (approximately 90%) were obtained from decaying mangrove leaves by the direct plating technique, while only 10% was obtained from seawater by the pine pollen baiting technique (data not shown). Twenty-six selected strains of different colony morphology were examined for microscopic morphological features and molecular phylogeny based on SSU rDNA sequences (Table 1).

The selected thraustochytrids can be categorized into seven major groups based on morphological features and molecular evidence (Table 2, Figures 2 and 3). The analyses of SSU rDNA sequences resulted in nine MPTs [tree length=2427 steps, CI=0.3976, RI=0.776, and RC=0.413], and the best tree topology as determined by the K-H test is presented in Figure 3. Of the 868 total characters, 360 were constant, 94 were parsimony-uninformative and 414 were parsimony-informative.

Figure 2:

Thraustochytrid colonies and cells.

(A–D) Aurantiochytrium limacinum (Group I, JS434). (E–H) Aurantiochytrium sp. (Group II, JS736). (I–L) Thraustochytrium sp. (Group III, JS489). (M–P) Parietichytrium sarkarianum (Group IV, JS510). (Q–X) Unidentified thraustochytrids (Group V, JS660). (Y–AB) Schizochytrium sp. (Group VI, JS1089). (AC–AJ) Thraustochytrium sp. (Group VII, JS974).

Figure 3:

One of the nine most parsimonious trees resulting from maximum parsimony analyses obtained from nuclear SSU rDNA sequence analyses of thraustochytrids.

Maximum parsimony (left) and maximum likelihood (right) bootstrap values greater than 50% are shown above each branch. Bayesian posterior probabilities greater than 0.95 are given below each branch. Scale bar indicates 10 character state changes.

Sequences of the isolated thraustochytrids cluster in the Thraustochytridae, Labyrinthulomycetes (Figure 3). Group I comprises seven thraustochytrid strains (JS446, JS989, JS828, JS988, JS434, JS827, JS490) possessing similar morphological features, except for JS490, which had smaller zoosporangia (Figure 2A–D). They are closely related to Aurantiochytrium limacinum (D. Honda et Yokochi) R. Yokoyama et D. Honda sequences retrieved from GenBank. Group II contains three strains (JS702, JS732, JS736) forming a strongly supported monophyletic group without any closely related taxa. They share similar morphological characteristics with the genus Aurantiochytrium R. Yokoyama et D. Honda, but differ from A. limacinum in colony morphology and growth rate (Figure 2E–H). Group III comprises seven strains (JS964, JS679, JS489, JS727, JS947, JS1085, JS1115), clustering with species of Thraustochytriidae and Thraustochytrium striatum Joa. Schneid., although JS1085 nestled separately from the others. The vegetative cell and zoosporangium measurement varied among strains within this group (Figure 2I–L). Moreover, two isolates (JS1087, JS510) nestled within Group IV and have a close affinity with strains of Parietichytrium sarkarianum (A. Gaertn.) R. Yokoyama, Salleh et D. Honda, with sequence similarities of 98.8–99.7%. Our strains share morphological features with P. sarkarianum in possessing vegetative globose cells, with thin walls. Amoeboid cells were actively motile and the cell wall was persistent after releasing the protoplast (Figure 2M–P). Four isolates (JS660, JS659, JS1093 and JS1103) are well placed in Group V without any closely related reference strains. The morphological characteristics for this group are unique (Figure 2Q–X). JS659 and JS660 possessed large vesicles inside the zoosporangia, whereas JS1093 and JS1103 possessed smaller vegetative cells and zoosporangia. Isolates JS1089 and JS1128 (Figure 2Y–AB) lie within Group VI comprising Schizochytrium aggregatum S. Goldst. et Belsky AB022106 and Schizochytrium sp. AB290576 with nucleotide similarity ranging from 92.9 to 97.2%. One thraustochytrid strain (JS974) forms a clade with T. gaertnerium R. Jain, Raghuk., Bongiorni et Aggarwal AY705753 with 88% sequence similarity (Group VII). Strain JS974 is distinct in morphological characterization. It had irregularly shaped colonies, thalli with thick hyaline walls, a globose sporangium and cytoplasm that cleaved into zoospores. Multiproliferous bodies (1–6 cells) were produced. Zoosporangia contained large vesicles with the protoplast retracted from the mature zoosporangium wall. Numerous cigar-shaped, limax amoebae were observed (Figure 2AC–AJ).

Polyunsaturated fatty acid composition

Mean values of cell biomass, TFA and PUFA composition of the isolated thraustochytrids are shown in Table 3 and Figure 4. Profiles of the fatty acids, particularly arachidonic acid (ARA, C20:4), eicosapentaenoic acid (EPA, C20:5), docosapentaenoic acid (DPA, C22:5) and docosahexaenoic acid (DHA, C22:6), were determined in each group. The production of C22 acids, particularly DHA (C22:6), ranged from 7 to 221 mg l⁻¹, whereas the C20 (C20:4 and C20:5) content ranged from 0.3 to 16.8 mg l⁻¹. The highest contents of DPA, DHA and EPA were obtained from Group II strains and the highest content of ARA was found in Group IV (Table 3).

Figure 4:

Polyunsaturated fatty acid profiles of the seven phylogenetic thraustochytrid groups isolated from the present study (as% of total fatty acids).

The fatty acid composition (as% of TFA) of individual strains is shown in Table 4. DHA was the predominant fatty acid present in all groups, as well as palmitic acid (C16:0) for Groups I and IV and pentadecanoic acid (C15:0) for Groups II and III. All groups studied contained high levels of C22 PUFAs, ranging from 34.2 to 61.4% of TFA, with lower levels of C20 PUFAs, varying from 1.8 to 20.5% of TFA. Group I contained the highest level of DHA (51.7% of TFA). Group V had the highest levels of ARA and EPA (8.2 and 12.4% of TFA, respectively). In addition, Group VII contained relatively high levels of DPA (27.3% of TFA; Table 4).

Cryopreservation of thraustochytrids

Effect of cryoprotective agents for cryopreservation at 80°C

The survival rates of thraustochytrids for all tested CPAs varied with species, whereas no viability of thraustochytrid cells was observed for the control experiment (no CPA added). Figure 5A shows the survival rates of each species after freezing at −80°C. Parietichytrium sarkarianum JS510 had a cell density before freezing of 7.3×10⁷ CFU ml⁻¹ and retained concentrations of 1.4×10⁷−6.0×10⁷ CFU ml⁻¹ (20–61% viability) after 6 months storage for all CPAs tested. Three CPAs (D, G and SG) provided the highest survival rates for JS510 after preservation for 6 months (p<0.05). Aurantiochytrium sp. JS702 had a cell density before freezing of 3.6×10⁶ CFU ml⁻¹ and retained concentrations of 1.1×10⁵–6.0×10⁶ CFU ml⁻¹ (3–29% viability) after 6 months of storage for all CPAs tested. Of these CPAs: D, HD and SG gave the lowest cell viability (p<0.05), while TG and G had the highest survival rates (29% and 16% viability, respectively) (p<0.05).

Figure 5:

Results of cryopreservation studies on Thai thraustochytrids.

(A) Mean viability (±standard deviation, n=15) of four strains of marine thraustochytrids after cryopreservation for 6 months at −80°C using five different cryoprotective agents (D=5% DMSO; G=10% glycerol; HD=30% horse serum combined with 10% DMSO; SG=17% skim milk combined with 20% glycerol; TG=5% trehalose combined with 10% glycerol). (B) Mean viability (±standard deviation, n=15) of four strains of marine thraustochytrids after cryopreservation for 12 months at −80°C and in vapor-phase nitrogen (LN) using the optimal cryoprotective agent for each strain (shown in brackets).

Thraustochytrium sp. JS1085 had a cell density before freezing of 2.6×10⁸ CFU ml⁻¹, and it retained concentrations of 1.7×10⁷−5.6×10⁷ CFU ml⁻¹ (6–21% viability) after 6 months of storage for all CPAs tested. All five CPAs yielded a similar range of survival rates, with no significant differences (p>0.05). Finally, Thraustochytrium sp. JS1089 had an initial cell density of 7.0×10⁶ CFU ml⁻¹ and retained concentrations of 3.2×10⁵−5.8×10⁶ CFU ml⁻¹ (4–60% viability) after 6 months of storage for all CPAs tested. The viability for JS1089 preserved with TG was the highest (60%) followed by G (17%).

Cryopreservation of selected thraustochytrids in different freezing conditions

The survival rates of thraustochytrids preserved at −80°C and the vapor-phase of nitrogen were determined using the best CPAs for each thraustochytrid tested over 6 months. The best CPA used for cryopreservation at −80°C and liquid nitrogen for JS510 was 10% glycerol (G), while for JS702, JS1085 and JS1089, 5% trehalose combined with 10% glycerol (TG) was the best. The survival rates of all strains tested at −80°C and in liquid nitrogen were similar during the first month of preservation (results not shown), but were significantly greater after freezing in liquid nitrogen for 12 months than those kept at −80°C (Figure 5B). Parietichytrium sarkarianum JS510 retained a cell concentration of 2.0×10⁷ CFU ml⁻¹ (20% viability) after freezing at −80°C, whereas the cell density after freezing in liquid nitrogen was 3.5×10⁷ CFU ml⁻¹ (35% viability). Aurantiochytrium sp. JS702 had a cell concentration of 7.0×10⁵ CFU ml⁻¹ (20% viability) after freezing at −80°C, but 1.5×10⁶ CFU ml⁻¹ (43% viability) after freezing in liquid nitrogen. Similarly, Thraustochytrium sp. JS1085 showed 11% viability (3.1×10⁷ CFU ml⁻¹) after freezing at −80°C, but 24% viability (6.5×10⁷ CFU ml⁻¹) after freezing in liquid nitrogen. Finally, Thraustochytrium sp. JS1089 showed 12% viability (8.7×10⁵ CFU ml⁻¹) after freezing at −80°C, but 21% viability (1.5×10⁶ CFU ml⁻¹) after freezing in liquid nitrogen.

Isolation, diversity and fatty acid production of marine thraustochytrids from Thailand

In this study, thraustochytrids isolated from decaying mangrove leaves represented the dominant group, whereas a small proportion of thraustochytrids was isolated from seawater using the pine pollen baiting technique (Porter 1990). Most of the strains isolated from pine pollen baiting were placed in clades II, V and VII in the phylogenetic tree (Figure 3). The direct plating of fallen mangrove leaves poses a higher risk of contamination by rapidly growing fungi, yeasts and antibiotic-resistant bacteria (Gupta et al. 2013). Pollen grains have been widely used as a bait for zoosporic fungi and actinobacteria, as well as thraustochytrids, as they were found to be a rich source of sugars, starch, protein, amino acids, lipid, sterol and other trace vitamins (Hayakawa et al. 1991, Phuphumirat et al. 2011, Gupta et al. 2013). In this study, pine pollen baiting was found to be a suitable technique for obtaining slow-growing thraustochytrids from seawater, especially those that potentially represent new lineages.

Seven major lineages of the isolated thraustochytrids can be distinguished based on morphological features and molecular phylogeny. Groups I and IV can be identified as Aurantiochytrium limacinum and Parietichytrium sarkarianum, respectively, based on their morphological features. Groups II, III, and VI can be identified to generic level as Aurantiochytrium sp., Schizochytrium sp. and Thraustochytrium sp., respectively. Finally, observations of the remaining taxa in Groups V and VII revealed that they could potentially represent new lineages within the Labyrinthulomycetes based on morphological and molecular evidence. Morphologically, all members nestled in Group V possessed hyaline thick-walled thalli and undeveloped ectoplasmic net elements. Strains JS659 and JS660 had large zoosporangia (20–177 μm), with large vesicles appearing inside the zoosporangia (Figure 2Q–X). In addition, strain JS974 (clade VII, Figure 2AC–AJ) was distinct in its irregularly shaped colonies with an undulating margin, sometimes producing elongated amorphous cells at the margin of the colony. The zoosporangia were unique with large vesicles and ectoplasmic net elements. Numerous cigar-shaped and limax amoebae were also observed. Comparison of Thraustochytrium sp. JS974 with T. gaertnerium, the most closely similar taxon, revealed that JS974 differs by possessing larger zoosporangia and a greater number of proliferation bodies (Bongiorni et al. 2005a).

This study assessed the diversity of newly isolated marine thraustochytrids from various mangrove locations in Thailand. The most common species was identified as A. limacinum, with other strains identified as species of the genera Schizochytrium, Parietichytrium and Thraustochytrium. This is in agreement with the observations of Yang et al. (2010), in which newly isolated thraustochytrids from Taiwan belonged to Aurantiochytrium and Thraustochytrium, although Aplanochytrium and Oblongichytrium R. Yokoyama et D. Honda were also found. Chang et al. (2012) identified four genera, namely Aurantiochytrium, Schizochytrium, Thraustochytrium and Ulkenia A. Gaertn. from Australian marine environments. Moreover, Gupta et al. (2013) identified thraustochytrid strains from Australia as species of Schizochytrium, Thraustochytrium and Ulkenia using a pine pollen baiting technique. In addition, our observation is in agreement with a recent study by Ou et al. (2016), in which Aurantiochytrium was dominant in Malaysian mangroves. Differences in species diversity and composition may be attributed to different substrata and environmental factors, such as water temperature, salinity and nutrient sources (Raghukumar 2002, Ueda et al. 2015). Identification of thraustochytrids has been found to be difficult. No single morphological character can be used to classify these organisms (Honda et al. 1999). Major limitations for the identification of marine thraustochytrids include their morphological complexity and the limited number of named sequences in the public databases for comparison.

In our study, all the thraustochytrids isolated from Thailand produced PUFAs, such as ARA, EPA, DPA and DHA. This result is similar to earlier studies by Chang et al. (2012) and Gupta et al. (2016), which showed the presence of PUFAs in thraustochytrids isolated from various habitats in Australia and India. They noted that DHA constituted 20–50% of TFA, and was the major PUFA found in their strains. The others, ranging from 1 to 12% of TFA, were DPA, ARA and EPA. In the present study, the highest yielding isolates of DHA appeared in Group I belonging to A. limacinum (51% of TFA), which is similar to the strains BL8 and BL10 (51% and 47%, respectively) reported by Chang et al. (2012). Since the biomass and TFA production of Group I isolates (1.0 g l⁻¹, 10.2% w/w) were lower than those of Group II isolates (3.1 g l⁻¹, 19.4% w/w), however, the DHA production of Group I isolates was lower than Group II per unit volume (52.7 and 221.9 mg l⁻¹, respectively). Thus, the strains in Groups I and II should be examined for further manipulation in order to maximize DHA production and evaluate their potential for commercial application. Thraustochytrids from other groups produced other C20–22 PUFAs including ARA, EPA and DPA ranging from 0.3 to 27.3% of TFA. Interestingly, EPA is an important fatty acid for commercial applications. The present study indicates that the level of EPA produced by the strains in this study (12.3% of TFA) is comparable with that of Australian strains (12.6%; Gupta et al. 2016). The results also showed that the highest amount of DPA (27.3% of TFA) was produced by a potentially new thraustochytrid lineage JS974 (Table 4).

PUFA composition is a key chemotaxonomic character for thraustochytrids, in addition to saturated fatty acids and the presence of odd chain PUFA (OC-PUFA; Huang et al. 2003, Yokoyama et al. 2007, Chang et al. 2014). The present study revealed that the profile and proportion of PUFAs (particularly C20–22) of the strains clustered in the same phylogenetic clade were similar (Figure 4, Tables 3 and 4). However, the fatty acid profile might be affected by culture conditions (Chang et al. 2014). Yokoyama et al. (2007) distinguished the genus-level phylogenetic groups in the Labyrinthulomycetes by a combination of morphological and chemotaxonomic features such as PUFA profiles and carotenoid pigments. Thus, it is necessary to establish taxonomic criteria for thraustochytrids based on a combination of phenotypic, chemical and molecular characteristics.

Cryopreservation of selected marine thraustochytrids

The survival rates of thraustochytrid strains isolated from Thailand after preservation in a −80°C freezer and vapor-phase nitrogen (−187°C) varied between species. Aurantiochytrium sp. JS702 and Thraustochytrium sp. JS1085 exhibited relatively poor survival rates after being frozen at −80°C. Parietichytrium sarkarianum JS510 and Schizochytrium sp. JS1089, with larger vegetative cells and zoosporangia, appeared to be more tolerant to cryopreservation. This is in agreement with a study suggesting that algae with a large cell size are more tolerant to cryopreservation (Day and Brand 2005). Prolonged storage at −80°C as well as in liquid nitrogen was found to reduce cell viability. The loss of viability is related to cell damage due to the formation of ice and osmotic pressures (Snell 1991, Miyamoto-Shinohara et al. 2000).

Our results have established suitable CPAs and cryopreservation conditions for the economically valuable marine thraustochytrids recently discovered in Thailand. The most effective CPAs after cryopreservation at −80°C were 10% glycerol (G) and the combination of 5% trehalose and 10% glycerol (TG), which maintained good viability for most of the thraustochytrid strains tested. Glycerol and trehalose have been widely used as CPAs in preserving various microorganisms. Glycerol has the ability to penetrate both cell wall and cell membrane, whereas trehalose is a non-permeable CPA (Hubálek 2003). The use of combined CPA types, which have synergistic effects, may enhance the viability of cells or tissues after cryopreservation (Crowe et al. 1984).

In the present study, thraustochytrids stored in vapor-phase nitrogen showed higher viability than preservation in −80°C (Figure 5B). After 12-month storage in liquid nitrogen, the survival rates of thraustochytrids ranged from 21 to 43% in different species. The method for long-term preservation of thraustochytrid cultures in liquid nitrogen was originally developed by Bremer (2000). The cultures were preserved under liquid nitrogen together with pollen grains, without the use of a cryoprotective agent. Some cultures survived after many years of storage, but not all (Bremer 2000). The only recent documented cryopreservation technique for marine thraustochytrids was developed by Cox et al. (2009). They noted that the use of a combination of 30% horse serum and 10% DMSO (“HD” in this study) was the most effective CPA for liquid nitrogen cryopreservation (1 month of storage) of thraustochytrids from New Zealand. Storage in liquid or vapor-phase nitrogen is the most universally applicable preservation method for various microbes, such as fungi, bacteria, viruses and protozoa, as well as animal, algal and plant cells (Snell 1991).

However, there are some limitations to using liquid nitrogen for long-term preservation of microbes. It requires specialized equipment, operating and maintenance systems and a regular supply of liquid nitrogen (Day and Brand 2005). Our observations revealed that the thraustochytrid strains tested can be frozen in vapor-phase liquid nitrogen and retain 26–50% viability for 6 months (data not shown). Nevertheless, the effectiveness of cryopreservation in microorganisms will be influenced by multiple factors, including species or strain, cell structure, growth phase and rate, growth medium, incubation, temperature, cell density before freezing, type and concentration of CPAs, cooling rate, warming rate, preserved time and recovery medium (Hubálek 2003).

In conclusion, our study has investigated the diversity of newly isolated marine thraustochytrids from Thailand. All of the strains isolated produced commercially useful high-value fatty acids. Two of the seven thraustochytrid groups isolated could represent potential new lineages in the Labyrinthulomycetes based on their unique morphological features, molecular phylogeny and fatty acid profiles. The thraustochytrids were successfully preserved at −80°C or in liquid nitrogen. Glycerol, alone and in combination with trehalose, yielded relatively high survival rates for the thraustochytrid strains. The thraustochytrids tested were well-preserved for at least 1 year in vapor-phase liquid nitrogen. The information provided in this study could be applied to long term preservation of the economically valuable marine Labyrinthulomycetes. Further investigation of media optimization for higher PUFA yields and production is necessary.