A Simple and Rapid Cryopreservation Technique for Ciliates: a Long-Term Storage Procedure Used for Marine Scuticociliates

Journal of Eukaryotic Microbiology ISSN 1066-5234

ORIGINAL ARTICLE A Simple and Rapid Cryopreservation Technique for Ciliates: A Long-Term Storage Procedure Used for Marine Scuticociliates

Yongqiang Liua,b, Bei Nana,b, Lili Duana,b, Ting Chenga,b, William A. Bourlandc, Mingjian Liua & Yan Zhaod a Institute of Evolution & Marine Biodiversity, Ocean University of China, Qingdao 266003, China b Laboratory for Marine Biology and Biotechnology, Qingdao National Laboratory for Marine Science and Technology, Qingdao 266003, China c Department of Biological Sciences, Boise State University, Boise, Idaho 83725-1515, USA d Research Centre for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China

Keywords ABSTRACT Freezing; liquid nitrogen; methodology; protozoa; recovery efﬁciency. Pseudocohnilembus persalinus is a free-living marine scuticociliate that, as a new model organism, has been used in a wide variety of studies. However, Correspondence long-term laboratory maintenance for this species is mainly achieved by sub- M. Liu, Institute of Evolution & Marine culture that requires rigorous culture environments and, too often, cultures of Biodiversity, Ocean University of China, the organism die out for a variety of reasons. Successful transport of viable Qingdao 266003, China cultures also poses problems for researchers. This study describes a simple Telephone number: +86 532 82031676; and rapid protocol for long-term cryopreservation of P. persalinus. The effects FAX number: +86 532 82031676; of physiological states of individuals before freezing, the type and concentra- e-mail: [email protected] tion of cryoprotectant, and optimal temperatures for freezing and thawing and were assessed. A cryopreservation protocol, using a mixture of 30% glycerol Y. Zhao, Research Center for Eco-Environ- and 70% concentrated P. persalinus cell culture, incorporating rate-controlled mental Sciences, Chinese Academy of freezing at 80 °C before liquid nitrogen storage, maintained a high recovery Sciences, Beijing 100085, China efﬁciency after 8 wk of storage. These results suggest that broader application Telephone number: +86 10 62841639; of this protocol to build a cryopreserved marine protozoa culture bank for bio- FAX number: +86 10 62841639; logical studies may be possible. e-mail: [email protected]

Received: 6 January 2019; revised 22 February 2019; accepted March 15, 2019. Early View publication 12 April, 2019 doi:10.1111/jeu.12730

CILIATES are ideal model organisms for biological research isolations of trophic cells into proteose peptone-based or and have already been used in various ﬁelds, such as genet- bacterized media such as wheatgrass medium, and incu- ics, ecology, epigenetics, developmental cellular biology, bation under controlled environments (Anderson et al. and ageing studies (Gao et al. 2016; Li et al. 2018; Wang 2009; Cassidy-Hanley 2012; Zheng et al. 2015). However, et al. 2017a,b; Yan et al. 2016; Zhao et al. 2018; Zheng the above-mentioned preservation methods are inconsis- et al. 2018). In recent years, the study of ciliate biodiversity, tently successful, often due to unwanted encystment of molecular phylogeny, and cell development has steadily many species at low food concentrations or under subopti- increased, highlighting the need for the improvement of cili- mal environmental conditions (Benca tova and Tirjakova ate preservation methods, especially cryopreservation 2018; Muller€ 2007; Tsutsumi et al. 2004). For instance, (Chen et al. 2017, 2018; Pan et al. 2017; Qu et al. 2018; Pseudocohnilembus persalinus can form cysts under con- Sheng et al. 2018; Zhang et al. 2018; Zhao et al. 2017). ditions of food deﬁciency, pH changes, low oxygen con- For most cultivatable ciliate species, to maintain use- centration, the accumulation of excretory products, able, stable populations generally involves repeated overcrowding, low temperatures, and dessication (Fenchel

1990; Tsutsumi et al. 2004). Many ciliate species are cap- micropipettes from raw cultures, washed 3 or 4 times, able of excystment, growth, and then re-encystment after and then transferred to Petri dishes with filtered, auto- being transferred to a fresh culture medium (Day et al. claved seawater at 25 °C. Rice grains were added to pro- 2007). However, for some species, problems with the mote the growth of bacteria, the food of the ciliates. long-term stability of the cysts held at “normal” environ- Living cells were observed, using bright field and differ- mental temperatures (5–20 °C) may decrease the percent- ential interference contrast microscopy at 100–1,0009 age of excystment over time (Muller€ et al. 2008). Thus, magnification. The ciliature and argyrome structures were there are still many practical difficulties in the preservation revealed, using the protargol, Chatton-Lwoff silver nitrate of cyst-forming ciliates due to low recovery efficiency. and the silver carbonate staining methods (Corliss 1953; Cryopreservation in liquid nitrogen seems to be an opti- Ma et al. 2003; Wilbert 1975). Measurements were made mal method for preservation of cyst-forming protozoa. It under 1,0009 magnification with a calibrated ocular has the potential to ensure long-term stability of these micrometer. organisms ex situ, and to maintain characteristic genetic traits since biological activities will be completely stopped Culture medium, mass cultivation, and induction of at the storage temperature (196 °C). Thus, cryopreserva- encystment tion could be a useful alternative or complement to long- term culture maintenance for ciliate research (Ashwood- Escherichia coli (HST04 dam/dcm Competent Cells, Smith and Farrant 1980; Mazur 1984). Previous studies of Code No. 9129; TaKaRa Co., Ltd., Beijing, China) cells protist cryopreservation have mainly focused on a broad were used as food for Pseudocohnilembus persalinus. diversity of rumen-parasitic and freshwater ciliates (Muller€ The E. coli were cultivated at 37 °C in a flask with Luria- et al. 2008). The low-temperature storage of Tetrahymena Bertani (LB) broth on a shaker at 200 rpm for 24 h and species is probably the most time-tested approach, and grown to an optical density of 2.0 at 600 nm (OD600). successful long-term storage of T. thermophila has been Then 25 ml of culture solution (OD600 = 2.0) was col- accomplished by this method (Cassidy-Hanley et al. 2010; lected and washed 2 or 3 times, using sterile filtered sea- Simon 1982). However, large-scale cultivation methods for water by centrifugation at 6,500 g for 5 min. Finally, 10 ml marine ciliate species are less dependable than for fresh- sterile filtered seawater was added to the bacterial sus- water species due to less well-defined living conditions pension which was stored at 4 °C for use as food for the (such as salinity), and the cryoprotectants that have been ciliates. used for freshwater species are sometimes harmful to Pseudocohnilembus persalinus cells were cultured at marine ciliates (Fuller 2004). 25 °C in a stationary incubator, and 10 ll of prepared bac- A practical method of cryopreservation and recovery for terial suspension was added every 6–8 h to 5 ml of the the marine ciliate Uronema marinum was reported by ciliate culture before freezing or after thawing. Encyst- Anderson et al. (2009), using a 2-ml system containing ment was induced by starving the P. persalinus cells 20% glycerol, 10% fetal bovine serum, and 70% U. mar- (2 days after initiation of culture, abundance about inum cells for cryopreservation. It allowed successful 4 9 104 cells/ml) in sterilized seawater for about 5 days. recovery of U. marinum. However, the recovery was time- consuming, required a high concentration of cells for cry- Immunofluorescence staining and imaging opreservation, and yielded a low survival abundance after recovery (Anderson et al. 2009; Cassidy-Hanley et al. Immuofluoresence (IF) staining was performed according 2010). to previous studies (Gao et al. 2013; Wang et al. 2017a,b). Based on previous studies, we developed a repeatable, The individuals of Pseudocohnilembus persalinus were effective, and user-friendly cryopreservation method for immunostained with an anti-acetylated a-tubulin (Lys40) the marine scuticociliate Pseudocohnilembus persalinus monoclonal antibody (Catalog: 32-2700, 1:500; Thermo that allowed recovery of viable cells after storage in liquid Scientific, Invitrogen, Shanghai, China) to label the cilia. nitrogen for up to 8 wk. The cells could be successfully Cells were incubated with the primary antibodies at 4 °C recovered with high efficiency using this protocol. The overnight and then were incubated in the secondary anti- results have enabled us to begin establishing cryobanks of body (Goat anti-Mouse Ig G1, Alexa Fluor 555, Catalog: A- free-living marine ciliates. 21127, 1:500; Thermo Scientific, Invitrogen, Maryland) at 25 °C for 1 h. Digital images were obtained by using a Motion Blitz EoSens Mini1 camera (Mikrotron, Unterschlei MATERIALS AND METHODS b heim, Germany) mounted on a Leica Sp8 confocal microscope with a 1009 oil immersion objective (numeri- Sampling, identification, and preliminary cultivation cal aperture is 1.40). Pseudocohnilembus persalinus was collected from surface seawater in Qingdao, northern China (36°03043″N; Swimming paths and relative swimming velocity 120°19012″E), on 3rd September 2016 at a water temperature of 24 °C and a salinity of 30&. Preliminary cultivation Swimming speed was calculated by measuring the length was performed according to previously reported methods of swim paths from digital videos with digital imaging soft- (Luo et al. 2018). Individual cells were picked with ware (ImageJ v1.6.0, Bethesda, Maryland). Digital videos

© 2019 International Society of Protistologists Journal of Eukaryotic Microbiology 2019, 66, 836–848 837 Cryopreservation for a Marine Ciliate Liu et al. of swimming cells were captured at a rate of 400 frames Table 1. The abundance of recovered cells was counted per second (fps), using an Olympus BX53F microscope five times (6, 12, 24, 36, and 48 h after thawing, using a with an Olympus DP74 camera. Image processing steps hemocytometer (Bright-line hemacytometer, Code No. were performed with Image J (National Institutes of 3200; Hausser scientific Co., Horsham, PA) under the Health, Bethesda, MD) to show the swimming path in half stereomicroscope at 11.59 magnification. Counts were a second. Relative swimming velocity was calculated by repeated four times per tube, and three tubes were dividing the swimming path length by the number of paths counted for each sample. Statistical analyses were per- in each field view (three replicates). formed using ORIGIN v 9.0 (OriginLab Co., Ltd., North- ampton, MA). Arithmetic means and standard deviations were calculated for the abundance of each group. Signifi- Rate-controlled freezing cant difference analysis was performed, using One-way The trophic cells used in this part of the study were in log ANOVA with LSD test on SPSS 19.0 (Norusis 2008). phase growth (1 9 105 cells/ml), after culturing at 25 °C for 3 days. Before freezing, the material was concentrated RESULTS to approximately 8 9 105 cells/ml at 2,000 g for 5 min at 25 °C (Table 1). Experimental replicates (n = 3) were asep- Morphology and relative swimming velocity of tically transferred into 1.8 ml cryotubes (NuncTM cryotubeTM Qingdao population of Pseudocohnilembus persalinus vials, Code No. 375418; Thermo Co., Ltd., Shanghai, China) containing 350 ll mixture of the inoculum and the The morphological characters of Pseudocohnilembus per- cryoprotectant [either glycerol or Dimethyl Sulfoxide salinus in vivo and after silver impregnation, corresponded (DMSO, Code No. D8418; Sigma-Aldrich, Shanghai, well with the original description and redescriptions (Kim China), concentration of cryoprotectant ranges from 10% et al. 2004; Pan et al. 2012; Song 2000) (Fig. 1A–G). The to 50%]. The cryotubes were sealed and gently inverted growth curve in Fig. 1H revealed that the population twice to ensure that all contents were completely mixed. reached its highest abundance (3 9 105 cells/ml) 4 or Samples were exposed to cryoprotectant at 25 °C for 5 days after culturing with added bacterial suspension. 30 min before freezing, and then placed in a rate-con- The Qingdao population of P. persalinus, like the original trolled cooler (Nalgene Mr. Frosty, Code No. 5100; population (Song 2000), shows a morphologic variability Thermo Co., Ltd.) and frozen from 25 °Cto80 °C at the under different nutritional conditions. The body size of indi- rate of 1.0 °C/min. viduals under adequate nutrition is about 25–35 lm 9 10– 15 lm in vivo, drop-shape in ventral view with the anterior end narrowed and the posterior end rounded, and the cell Recovery assessment appears dark due to numerous food vacuoles (Fig. 2A). The abundance of recovered cells was calculated after a When in low nutrition conditions, the body is narrower, 3-day storage in liquid nitrogen or at 80 °C (except in about 25–35 lm 9 8–10 lm in vivo, and the cells are Experiment 4). Detailed recovery steps are shown in more transparent owing to the fewer food vacuoles

Table 1. Freezing and recovery protocol

Freezing Step 1 Add 10 ll prepared bacterial suspension to 5 ml cell culture every 6–8 h, and culture Pseudocohnilembus persalinus cells to log phase (1 9 105 cells/ml) at 25 °C in stationary incubator Step 2 Centrifuge cells at 2,000 g for 5 min at 25 °C, then remove the supernatant (approximately 8 9 105 cells/ml) Step 3 Add cryoprotectant in the centrifuge tube and mix with the cells to obtain ﬁnal concentrations from 5% to 50%. Transfer 350 ll of the mixed sample to each pre-labeled cryotube Step 4 Place all cryotubes in the incubator at 25°C for 30 min Step 5 Put all cryotubes in a rate-controlled cooler (Nalgene Mr. Frosty, Code No. 5100; Thermo Co, Ltd., Shanghai, China) and freeze cells for 24 h at 1 °C/min Step 6 Transfer some cryotubes to liquid nitrogen and maintain other cryotubes in the rate-controlled cooler at 80 °C, then continue cryopreservation for at least 3 days Recovery Step 1 Heat 10 ml sterilized seawater in 15 ml centrifuge tube in water bath to desired temperature Step 2 Transfer cryotubes with cells to the beaker water bath for thawing. Then add 350 ll of heated sterilized seawater into the cryotubes to accelerate the thawing process Step 3 When thawing is complete, transfer contents of each cryotube (0.7 ml) to 5 ml sterilized seawater (25 °C) in cell culture plate and mix them well Step 4a Place the cell culture plate in a stationary incubator at 25 °C, and add 10 ll prepared bacterial suspension into the 5 ml cell culture every 6–8 h. Cultures are generally established within 24 h aAccording to the recovery results in experiment 4, the abundance of recovered cells after dilution is about 1.4 9 104 cells/ml (Table S3).

Figure 1 Pseudocohnilembus persalinus in vivo (A–D) and after staining (E–G), and time course of cell abundance in culture over 10 days (H). A– D, Right ventrolateral (A–C) and left ventrolateral (D, from Song et al. 2009) views of representative individuals. E, F, Ventral (E) and dorsal (F) views after protargol staining. Photos were color inverted using Photoshop. G, Immunoﬂuorescence staining of cells and cysts with an anti-acetylated a-tubulin antibody (red) and DAPI (blue). H, The starting concentration for the experiment shown in Fig. 1H is approximately 0.1 9 104 cells/ ml. Cell abundances were counted by hemocytometer every 6 h after inoculation with E. coli. Abundance reached a maximum (about 3 9 105 cells/ml) at 5 days and then entered a stationary period for about 2 days. Values in H are presented as arithmetic mean standard deviations (SD) of three replicates. Scale bars: 10 lm.

(Fig. 2B). As shown in Figure. 1G, cilia quickly degener- after thawing, while it took 24 h for encysted cells ated and then cysts formed in low-nutrient conditions. (Table 2). The time to reach countable abundance was Instead of a quick spiralling movement, some of the 48 h for both physiological states, but there was a signifi- growing cells under adequate nutrition rotated slowly in a cantly higher abundance for log-phase trophic cells frozen very limited space (Fig. 2C) when compared with starving with 10% DMSO in liquid nitrogen (Fig. 3). individuals (Fig. 2D). The swimming velocity of growing When using glycerol as the cryoprotectant, the time to cells was relatively reduced compared with starving ones the appearance of motile cells was similar for both physio- (Fig. 2E). logical states. However, log-phase trophic cells reached countable abundance earlier (Table 2). The log-phase trophic cells also reached higher abundances than the Experiment 1: Effects of different physiological states cysts did after thawing and culturing for 48 h (Fig. 3). of ciliates before freezing, types of cryoprotectants, Tests on different physiological states of the cells showed and varying cryoprotectant concentrations that the recovery efficiencies of log-phase trophic cells Two physiological states of the cells (log-phase trophic were higher than those of the cysts. cells vs. cysts) were tested, and two agents (DMSO and glycerol) were tested as cryoprotectants at three concen- Effect of different concentrations and types of trations (0%, 5%, 10%). The DMSO and glycerol treated cryoprotectants for log-phase trophic cells of cells were stored at either 80 °C or in liquid nitrogen Pseudocohnilembus persalinus (196 °C). Thawing was done in a water bath at 42 °C for When glycerol is used as cryoprotectant, the time to 1 min 30 s (Tables 1 and 2). The time to the appearance the appearance of motile ciliates was shorter than that of motile cells and the time to reach abundances count- for cells frozen with DMSO (about 2–4 h for glycerol able in the hemocytometer were compared. vs. 12 h for DMSO) (Table 2). Furthermore, after thawing and culturing for 24 h to 36 h, glycerol-preserved Effects of different physiological states of the cells cells achieved countable abundances faster than those before freezing preserved in DMSO (Table 2, Fig. 3). The number of When using DMSO as the cryoprotectant, for log-phase viable cells counted 48 h after thawing and the corre- trophic cells, the motile ciliates can be found within 12 h sponding abundances were significantly higher for

Figure 2 Photomicrographs of the appearances (A, B) and the swimming paths (C, D)ofPseudocohnilembus persalinus under growing (A, C) and starvation (B, D) conditions. Relative swimming velocities of growing and starving cells were calculated (E). A, B, The cells formed a dark appearance with adequate nutrition (A), while they became transparent when starving (B). C, D, The swimming paths of growing (C) and starving cells (D) in half a second, shown that growing individuals rotated slowly in a very limited space (C) compared to the starving cells (D). E, The motility of growing cells was relatively reduced and the corresponding data is presented as arithmetic mean standard deviations (SD). Student’s t-tests were used to identify statistically significant differences between starvation and growing cells relative velocity, **P < 0.01. glycerol-preserved cells than those in DMSO at both concentrations of glycerol (40% and 50%) did not greatly storage temperatures. These results indicated that glyc- increase the abundances of recovered cells and caused a erol is a more effective cryoprotectant than DMSO, and delayed achievement of countable abundances compared among the three tested concentrations (0%, 5%, 10%), to 20% or 30% glycerine (Fig. 4A). Abundances reached a 10% glycerol shows the optimal recovery efficiencies maximum when 30% glycerol was used. In general, the for log-phase trophic cells of Pseudocohnilembus abundances of cells thawed from liquid nitrogen storage persalinus. were slightly higher than those from maintenance at 80 °C (Fig. 4A). Experiment 2: Optimal concentration of glycerol for cryopreservation in Pseudocohnilembus persalinus Experiment 3: Effect of thawing temperature on the recovery of log-phase trophic cells in After determining glycerol to be superior to DMSO as cry- Pseudocohnilembus persalinus oprotectant, log-phase trophic cells were tested in five concentrations of glycerol (10%, 20%, 30%, 40%, and Log-phase trophic cells stored at 80 °C and in liquid 50%) at two different storage temperatures (80 °C and nitrogen (196 °C) for 3 days with 30% glycerol as the liquid nitrogen) for three days of cryopreservation cryoprotectant were used to determine the effect of thaw- (Table 3). The abundances of recovered cells were calcu- ing temperature on the recovery of Pseudocohnilembus lated every 6 h during 48 h of culturing (the first time persalinus. The trophic cells were thawed at five different were calculated at 6 h). Results were shown in Table S1. thawing temperatures (Table 4), and the abundances of Concentrations of glycerol between 10% and 30% pro- recovered cells were calculated every 12 h during 48 h of vided significantly higher numbers of recovered cells of culturing (the first time were calculated at 6 h). Results Pseudocohnilembus persalinus 48 h after thawing. Higher are shown in Table S2.

© 2019 International Society of Protistologists 840 Journal of Eukaryotic Microbiology 2019, 66, 836–848 i tal. et Liu ora fEkroi Microbiology Eukaryotic of Journal © 09ItrainlSceyo Protistologists of Society International 2019 2019, Table 2. Effect of physiological status and types of cryoprotectants on cell recovery efﬁciency 66

836–848 , Parametersa Effects of physiological status of the species and types of cryoprotectants

Life stage Trophic cells Cysts

Duration 3 days of freezing Thawing 42 °C water bath for 1 min 30 s Cryoprotectant DMSO Glycerol DMSO Glycerol

Concentration 0% 5% 10% 0% 5% 10% 0% 5% 10% 0% 5% 10% Pre-freezing Normal Normal Normal Normal Normal Slower Normal Normal Normal Normal Normal Normal motion swimming swimming swimming swimming swimming swimming swimming swimming swimming swimming swimming swimming status Slow freeze 80 °CLN80 °CLN 80 °CLN 80 °CLN80 °CLN 80 °CLN 80 °CLN 80 °CLN 80 °CLN 80 °CLN 80 °CLN 80 °CLN and store Time to ––12 h 12 h 12 h 12 h ––4h 4h 2h 2h 24h 24h 24h 24h 24h 24h 24h 24h 3h 3h 3h 3h motilityb Incubation periodc 6h ––––––––+ + ++ ++ ––––––––++++ 12 h ––++++––++ ++ +++ +++ ––––––––+ + ++ ++ 24 h ––++ ++ ++ ++ ––CCCC+ + ++ + + + + + +++ +++ +++ +++ 36 h ––+++ +++ +++ +++ ––CCCC++ ++ +++ +++ ++ ++ ++ ++ CCCC 48 h ––CCCC––CCCC+++ +++ CCCC+++ +++ CCCC

a 2 3 LN, liquid nitrogen; cell density: –, no moving cells; +, < 10 cells; ++, ≥ 10 cells; +++, ≥ 10 cells; C, countable using hemocytometer (≥ 10 cells). Ciliate Marine a for Cryopreservation bThe time to observe the ﬁrst moving cell after thawing. cThe result showed that trophic cells are superior to cysts and glycerol is superior to DMSO; for statistical comparison, see Fig. 3. 841 Cryopreservation for a Marine Ciliate Liu et al.

Figure 3 Abundances of recovered Pseudocohnilembus persalinus cells in different physiological states (log-phase trophic cells and cysts) exposed to different cryoprotectants (glycerol and DMSO) with different ﬁnal concentrations (5%, 10%). All cells were stored at 80 °C or in liquid nitrogen for 3 days and recovered at thaw temperature of 42 °C. The graph shows statistical results after thawing for 36 and 48 h (Experi- ment 1). Values are given as arithmetic mean standard deviations (SD) of three replicates. Student’s t-tests were used to identify statistically signiﬁcant differences between glycerol and DMSO, ***P < 0.001, **P < 0.01, *P < 0.05.

Table 3. Effect of glycerol concentrations and two freezing methods on cell recovery efﬁciency

Parametersa Effects of glycerol concentrations and two freezing methods

Life stage Trophic cells Duration of freezing 3 days Thawing 42 °C water bath for 1 min 30 s Cryoprotectant Glycerol Concentration 10% 20% 30% 40% 50% Slow freeze and store 80 °CLN 80 °CLN 80 °CLN 80 °CLN 80 °CLN Time to motilityb 2 h 2 h 0.5 h 0.5 h 0 h 0 h 5 h 5 h 6 h 6 h Incubation periodc 6h +++ +++ CCCC++ ++ + + 12 h C C C C C C ++ ++ ++ ++ 24 h C C C C C C C C +++ +++ 36h C C C C C CC CC C 48h C C C C C CC CC C aLN, liquid nitrogen; cell density: +, < 10 cells; ++, ≥ 10 cells; +++, ≥ 102 cells; C, countable using hemocytometer (≥ 103 cells). bThe time to observe the ﬁrst moving cell after thawing. cThe result showed that 30% glycerol is the optimum concentration; for statistical comparison, see Fig. 4A.

In general, the abundances of cells thawed from liquid culturing, the highest abundance of recovered cells was nitrogen storage (196 °C) were slightly higher at each obtained after thawing for 1 min 30 s at 42 °C (4.0 9 time point than those maintained at 80 °C. After 48 h of 104 cells/ml from liquid nitrogen and 2.5 9 104 cells/ml

© 2019 International Society of Protistologists 842 Journal of Eukaryotic Microbiology 2019, 66, 836–848 Liu et al. Cryopreservation for a Marine Ciliate from 80 °C at 48 h), and thawing for 1 min 20 s at liquids, protecting vital cell structures, and promoting vitri- 55 °C from liquid nitrogen also yielded a high abundance fication rather than catastrophic intracellular ice crystal for- of recovered cells at 48 h (Fig. 4B). The optimal thaw- mation (Gao and Critser 2000; Wolfe and Bryant 2001). ing temperature of 42 °C was used in the following However, in the case of encysted cells, the cyst wall may experiment. prevent or slow the access of the cryoprotectant to intracellular structures, allowing irreparable damage to at least some cells and preventing or decreasing cell recovery Experiment 4: Effect of freezing duration and storage (Muller€ et al. 2008; Pan et al. 2012). temperature on the recovery of log-phase trophic Pseudocohnilembus persalinus cells Effect of cryoprotectants To determine the effect of freezing duration on recovery efficiency, 30% glycerol was used as cryoprotectant for Although the mechanism of action of cryoprotectants is log-phase trophic cells, and the mixtures were cryopre- not fully understood, most cryoprotectants interact with served, at either 80 °C or in liquid nitrogen, for 2, 4, 6 or the phospholipids in the cell membrane, and the ability to 8 wk (Table 5). The results are shown in Table S3. permeate the cell membrane appears to be crucial. The After 2 wk of storage, the abundances of cells thawed functions of cryoprotectants include stabilizing the cell from liquid nitrogen (196 °C) were slightly higher than membrane, lessening the effects of dehydration on the those samples thawed from 80 °C. As the freezing time cells, changing the freezing point of cellular and extracellu- increased to four, six, and 8 wk, the abundance of cells lar solutions, and promoting vitrification to avoid ice crystal from liquid nitrogen remained similar to those frozen for formation in the intracellular fluids (Gao and Critser 2000; 2 wk, however, the abundance of cells thawed from stor- Wolfe and Bryant 2001). However, cryoprotectants can age at 80 °C tended to decline with longer freezing dura- also damage cells either directly by chemical toxicity or tion (Fig. 5). Additionally, in all cell cultures thawed from indirectly by affecting the intracellular and extracellular liquid nitrogen, swimming individuals could be observed osmotic pressures (Fahy 1986). Hence, toxicity and per- almost immediately after thawing while, in comparison, meability are both important in the choice of cryoprotec- attainment of motility was delayed in cells thawed from tants when developing a freezing protocol for a particular 80 °C (Table 5). According to our long-term experiments organism. and observations in the later period, the growth perfor- Both glycerol and DMSO have been successfully used mance and swimming behaviour of the post-freezing to preserve freshwater protozoa, rumen ciliates, and ani- Pseudocohnilembus persalinus was almost identical to mal cells (Fuller 2004; Hubalek 2003; Matsuo 2007; those of non-cryopreserved cultures for a period of time. Miyake et al. 2004). According to Lovelock’s theory of the actions of cryoprotectants, the increased salt concentrations (especially sodium chloride, the main con- DISCUSSION stituent of salts in marine water) would be ameliorated by the presence of the glycerol when cells are frozen Effect of pre-freezing physiological state of marine (Lovelock 1953). Thus, glycerol can be considered a suit- ciliates on recovery after cryopreservation able cryoprotectant for marine ciliates as it inhibits intra- Marine ciliates need specific environments for survival and cellular ice crystal formation (Fuller 2004; Lovelock tend to form cysts when subjected to adverse conditions 1954). However, when DMSO was used as a cryopro- (Foissner et al. 2008; Reid 1987). For example, their opti- tectant for the marine ciliate, Uronema marinum, all mal salinity should be kept between 10& and 30& and U. marinum samples frozen with DMSO (5–20%) were the temperature should be 20–25 °C with a sufficient food dead or burst after thawing, possibly due to higher toxi- supply (Fenchel 1990). It is common practice to cryopre- city of DMSO in marine ciliates compared with freshwa- serve any cell type at log-phase growth to yield optimal ter ciliates (e.g., Tetrahymena thermophila) (Anderson survival with traditional slow freezing protocols. In the et al. 2009). study of cryopreservation of Tetrahymena species, cells were starved in non-nutritive medium (10 mM Tris, pH Effect of thawing temperature 7.4) before freezing, which greatly enhanced the propor- tion of viable cells recovered (Cassidy-Hanley et al. 2010). Higher thawing temperature is thought to cause less cell The present results reveal that the physiological state of damage since any intracellular ice crystals would melt fas- cells prior to freezing is the most significant factor influ- ter (Mazur 1984; McGrath et al. 1975). However, exces- encing ciliate viability. In Experiment 1, the abundances of sive thawing temperature may also lead to cell damage in recovered cells were significantly higher for cells frozen in various ways, such as activation of oxygen radicals and/or log-phase than for those frozen in the encysted state. This decreasing the efficiency of enzymatic antioxidant path- seems counterintuitive since ciliates form resting cysts to ways (Alvarez and Storey 1992; Calamera et al. 2010). To maintain cell viability under adverse environmental condi- date, most cryopreservation protocols use a thawing tem- tions (Gutierrez et al. 1990, 2001). The cryoprotectant perature of 37 °C. In this study, Pseudocohnilembus per- enters unencysted cells through the cell membranes, low- salinus recovery was optimal with a thawing temperature ering the freezing point of extra- and intracellular biological of 42 °C for 1 min 30 s. Higher thawing temperatures,

Figure 4 Abundances of recovered cells after thawing and culturing for 48 h. A, Abundances of recovered cells exposed to different concentrations (10%, 20%, 30%, 40%, 50%) of glycerol. All cell samples were recovered at thawing temperature of 42 °C (Experiment 2). B, Abundances of recovered cells frozen with 30% glycerol and thawed at different temperatures (25 °C, 37 °C, 42 °C, 55 °C and 65 °C) (Experiment 3). Cells were stored at 80 °C or in liquid nitrogen for 3 days. The graph shows statistical results at 6, 12, 24, 36 and 48 h of culturing after thawing. Values are given as arithmetic mean standard deviations (SD) of three replicates. Student’s t-test, ***P < 0.001, **P < 0.01, *P < 0.05, #p no significant difference (see Table S1 and S2 for detailed results of significant difference analyses). even for shorter periods of time (65 °C for 50 s), showed (196 °C) for up to 8 wks have successfully been recov- a decline in the recovery efficiency. This situation is similar ered with high efficiency, indicating that cryopreservation to that found in Tetrahymena species (Cassidy-Hanley over the long term is possible in liquid nitrogen (Matsuo et al. 2010). Hence, thawing temperature should be 2007). strictly controlled to avoid potential cell damages. A tem- According to previous reports, various cell lines have perature of 42 °C for 1 min 30 s is recommended for cells been stored at 70 °Cto90 °C and successfully recov- thawed from liquid nitrogen because it yielded the highest ered (Bjornsson and Huebner 2002; Corsini and Mann abundances. 2005; Mazur 1984). Biochemical activities of ciliates are not totally stopped at such temperatures, and cellular damage and accumulation of toxic metabolites shorten Effect of the freezing duration and storage the time for which cells remain viable during storage (Ash- temperature wood-Smith and Farrant 1980; Gao and Critser 2000; The freezing duration and storage temperature are also Mazur 1984). By contrast, biochemical reactions (e.g., critical for successful long-term cryopreservation. In our enzyme activities) are completely stopped below 130 °C experiment, cells that cryopreserved in liquid nitrogen (glass transition temperature) when liquid water does not

Table 4. Effect of thawing temperature on cell recovery efﬁciency

Parametersa Effects of thawing temperature

Life stage Trophic cells Cryoprotectant Glycerol (30%) Duration of freezing 3 days Thaw temperature 25 °C37°C42°C55°C65°C Thaw time 1 min 50 s 1 min 40 s 1 min 30 s 1 min 20 s 50 s Slow freeze and store 80 °CLN 80 °CLN 80 °CLN 80 °CLN 80 °CLN Time to motilityb 2 h 0.5 h 0.5 h 0.5 h 0 h 0 h 2 h 0.5 h 0.5 h 0.5 h Incubation periodc 6h +++ C +++ C +++ C +++ C +++ C 12h C CC CC CC CC C 24h C CC CC CC CC C 36h C CC CC CC CC C 48h C CC CC CC CC C aLN, liquid nitrogen; cell density: +++, ≥ 102 cells; C, countable using hemocytometer (≥ 103 cells). bThe time to observe the ﬁrst moving cell after thawing. cThe result showed that liquid nitrogen is superior to 80 °C, and 42 °C is the optimum thawing temperature; for statistical comparison, see Fig. 4B.

Table 5. Effect of the freezing duration and storage temperature on cell recovery efﬁciency

Parametersa Effects of the freezing duration and storage temperature

Life stage Trophic cells Cryoprotectant Glycerol (30%) Thawing 42 °C water bath for 1 min 30 s Freezing duration 2 weeks 4 weeks 6 weeks 8 weeks Slow freeze and store 80 °CLN 80 °CLN 80 °CLN 80 °CLN Time to motilityb 3h 0h 4h 0h 5h 0h 6h 0h Incubation periodc 6h +++ C +++ C ++ C + C 12 h C C C C +++ C ++ C 24h C CC CC C+++ C 36h C CC CC CC C 48h C CC CC CC C aLN, liquid nitrogen; cell density: +, < 10 cells; ++, ≥ 10 cells; +++, ≥ 102 cells; C, countable using hemocytometer (≥ 103 cells). bThe time to observe the ﬁrst moving cell after thawing. cThe result showed that liquid nitrogen is superior to 80 °C, and the cell recovery efﬁciency remains stable when the cells were frozen in liquid nitrogen for 2 to 8 weeks; for statistical comparison, see Fig. 5.

exist and molecular movements are markedly slowed cryoprotectant, the rate of freezing, the storage tempera- (Ashwood-Smith and Farrant 1980; Gao and Critser 2000; ture, and the thawing temperature are crucial factors for a de Paoli 2005; Samarawickrema et al. 2001). It could successful cryopreservation of the marine ciliate Pseudo- explain our finding that the abundances of recovered cohnilembus persalinus. Our results indicate that using a Pseudocohnilembus persalinus cells thawed from liquid cryopreservation solution of 30% glycerol with 70% sus- nitrogen (196 °C) were greater than for those thawed pension of log-phase cells (v/v) and rate-controlled freezing from 80 °C for each of the four different freezing to 80 °C prior to storage in liquid nitrogen provides the durations. best long-term cryopreservation and optimal recovery efficiency. In addition, it is demonstrated that individuals of P. persalinus stored in liquid nitrogen for up to 8 wk could Summary of the protocol successfully be recovered, using this protocol. These find- The results of this study show that the growth phase of ings should encourage trials of this method for a broader the cell cultures, the type and concentration of range of marine ciliates and other protists.

Figure 5 Abundances of recovered trophic cells stored at 80 °C or in liquid nitrogen for 2, 4, 6 and 8 wk, respectively. All cells were recovered at 42 °C thawing temperature. Values are given as arithmetic mean standard deviations (SD) of three replicates. Student’s t-tests were used to identify statistically signiﬁcant differences between two storage temperatures, ***P < 0.001, **P < 0.01 (see Table S3 for the result of t-test among different freezing durations).

Uronema marinum isolated from farmed New Zealand groper ACKNOWLEDGMENTS (Polyprion oxygeneios). J. Microbiol. Methods, This work was supported by the National Natural Science 79:62–66. Foundation of China (project numbers: 31702009, Ashwood-Smith, M. & Farrant, J. 1980. Low temperature preser- 31801955), Natural Science Foundation of Shandong Pro- vation of cells, tissues and organs. Low temperature preservation in medicine and biology. University Park Press, Pitman vince (JQ201706), and the Fundamental Research Funds Medical Limited, Kent, England. p. 19–44. for the Central Universities (201841005). Many thanks to Benca tova, S. & Tirjakova, E. 2018. Light microscopy observations Ms. Yuanyuan Wang (Ocean University of China, OUC) for on the encystation and excystation processes of the ciliate Pha- assistance with the experiments, to Mr. Haibo Xie (OUC) codinium metchnikoffi (Ciliophora, Phacodiniidae), including for providing assistance with immunofluorescence micro- additional information on its resting cysts structure. Biologia, scopy imaging, to Ms. Deting Meng and Ms. Qin Ren 73:467–476. (OUC) for their assistance with data analysis, and to Prof. Bjornsson, C. S. & Huebner, E. 2002. Storage of various cell lines Shan Gao and Prof. Weibo Song (OUC) for their kind sug- at 70°Cor80°C in multi-well plates while attached to the – gestions during preparation of the manuscript. substratum. Biotechniques, 33:42 46. Calamera, J. C., Buffone, M. G., Doncel, G. F., Brugo-Olmedo, S., de Vincentiis, S., Calamera, M. M., Storey, B. T. & Alvarez, J. G. LITERATURE CITED 2010. Effect of thawing temperature on the motility recovery of cryopreserved human spermatozoa. Fertil. Steril., 93:789–794. Alvarez, J. G. & Storey, B. T. 1992. Evidence for increased lipid Cassidy-Hanley, D. M. 2012. Tetrahymena in the laboratory: Strain peroxidative damage and loss of superoxide dismutase activity resources, methods for culture, maintenance, and storage. as a mode of sublethal cryodamage to human sperm during cry- Methods Cell Biol., 109:237–276. opreservation. J. Androl., 13:232–241. Cassidy-Hanley, D. M., Smith, H. R. & Bruns, P. J. 2010. A sim- Anderson, S. A., Hulston, D. A., McVeagh, S. M., Webb, V. L. ple, efficient technique for freezing Tetrahymena thermophila. & Smith, P. J. 2009. In vitro culture and cryopreservation of J. Eukaryotic. Microbiol., 42:510–515.

Chen, X., Wang, Y. R., Sheng, Y. L., Warren, A. & Gao, S. 2018. Lovelock, J. E. 1954. The protective action of neutral solutes GPSit: An automated method for evolutionary analysis of non- against haemolysis by freezing and thawing. Biochem. J., culturable ciliated microeukaryotes. Mol. Ecol. Res., 18:700– 56:265–270. 713. Luo, X. T., Yan, Y., Shao, C., Al-Farraj, S. A., Bourland, W. & Chen, L. Y., Zhao, X. L., Shao, C., Miao, M. & Clamp, J. 2017. Song, W. B. 2018. Morphological, ontogenetic and molecular Morphology and phylogeny of two new ciliates, Sterkiella sinica data support strongylidiids as being closely related to Dorso- sp. nov. and Rubrioxytricha tsinlingensis sp. nov. (Protozoa, Cil- marginalia (Protozoa, Ciliophora) and reactivation of the family iophora, Hypotrichia) from north-west China. Syst. Biodivers., Strongylidiidae Faure-Fremiet, 1961. Zool. J. Linn. Soc., 15:131–142. 184:237–254. Corliss, J. O. 1953. Silver impregnation of ciliated protozoa by the Ma, H. G., Choi, J. K. & Song, W. B. 2003. An improved silver Chatton-Lwoff technic. Biotech. Histochem., 28:97–100. carbonate impregnation for marine ciliated protozoa. Acta Proto- Corsini, J. & Mann, E. 2005. Rapid cryopreservation of five mam- zool., 42:161–164. malian and one mosquito cell line at 80 °C while attached to Matsuo, J. 2007. A simple and rapid method for cryopreservation flasks in a serum free cryopreservative. Biol. Proced. Online, of Trichomonas vaginalis. Parasitol. Res., 101:907–911. 7:26–30. Mazur, P. 1984. Freezing of living cells: mechanisms and implica- Day, J. G., Achilles-Day, U., Brown, S. & Warren, A. 2007. Cultivation tions. Am. J. Physiol., 247:125–142. of algae and protozoa. In: Hurst, C., Crawford, R., Garland, J., Lip- McGrath, J. J., Cravalho, E. G. & Huggins, C. E. 1975. An experi- son, D., Mills, A. & Stetzenbach, L. (eds), Manual of Environmental mental comparison of intracellular ice formation and freeze- Microbiology, 3rd ed. ASM Press, Washington, DC. p. thaw survival of HeLa S-3 cells. Cryobiology, 12:540–550. 79–92. Miyake, Y., Karanis, P. & Uga, S. 2004. Cryopreservation of proto- Fahy, G. M. 1986. The relevance of cryoprotectant “toxicity” to zoan parasites. Cryobiology, 48:1–7. cryobiology. Cryobiology, 23:407–426. Muller,€ H. 2007. Live observation of excystment in the spirotrich Fenchel, T. 1990. Adaptive significance of polymorphic life cycles ciliate Meseres corlissi. Eur. J. Protistol., 43:95–100. in Protozoa: responses to starvation and refeeding in two spe- Muller,€ H., Achilles-Day, U. E. & Day, J. G. 2008. Cryopreserva- cies of marine ciliates. J. Exp. Mar. Biol. Ecol., 136: tion of resting cysts of the freshwater ciliate Meseres corlissi 159–177. by conventional two-step methods and one-step vitrification Foissner, W., Chao, A. & Katz, L. A. 2008. Diversity and geo- protocols. CryoLetters, 29:329–338. graphic distribution of ciliates (Protista: Ciliophora). Biodivers. Norusis, M. 2008. SPSS 16.0 Advanced Statistical Procedures Conserv., 17:345–363. Companion. Prentice Hall Press, Upper Saddle River, NJ. Fuller, B. J. 2004. Cryoprotectants: the essential antifreezes to Pan, H. B., Jiang, J. M., Fan, X. P., Al-Farraj, S. A. & Gao, S. protect life in the frozen state. CryoLetters, 25:375–388. 2017. Phylogeny and taxonomy of five poorly known species of Gao, F., Alan, W., Zhang, Q. Q., Gong, J., Miao, M., Sun, P., Xu, D. cyrtophorian ciliates (Protozoa, Ciliophora, Phyllopharyngea) P., Huang, J., Yi, Z. Z. & Song, W. B. 2016. The all-data-based from China seas. Zool. J. Linn. Soc., 180:475–492. evolutionary hypothesis of ciliated protists with a revised classifi- Pan, X. M., Ma, H. G., Shao, C., Lin, X. F. & Hu, X. Z. 2012. cation of the Phylum Ciliophora (Eukaryota, Alveolata). Sci. Rep., Stomatogenesis and morphological redescription of Pseudo- 6:24874. cohnilembus persalinus (Ciliophora: Scuticociliatida). Acta Hydro- Gao, D. & Critser, J. K. 2000. Mechanisms of cryoinjury in living biol. Sin., 36:489–494. cells. ILAR J., 41:187–196. de Paoli, P. 2005. Biobanking in microbiology: from sample collec- Gao, S., Xiong, J., Zhang, C. C., Berquist, B. R., Yang, R. D. Zhao, tion to epidemiology, diagnosis and research. FEMS Microbiol. M., Molascon, A. J., Kwiatkowski, S. Y., Yuan, D. X., Qin, Z. Rev., 29:897–910. H., Wen, J. F., Kapler, G. M., Andrews, P. C., Miao, W. & Liu, Qu, Z. S., Li, L. F., Lin, X. F., Stoeck, T., Pan, H. B., Al-Rasheid, Y. F. 2013. Impaired replication elongation in Tetrahymena K. A. S. & Song, W. B. 2018. Diversity of the cyrtophorid genus mutants deficient in histone H3 Lys 27 monomethylation. Chlamydodon (Protista, Ciliophora): its systematics and geo- Genes Dev., 27:1662–1679. graphic distribution, with taxonomic descriptions of three spe- Gutierrez, J. C., Callejas, S., Ben€ıtez, S. B. L., Martin-Gonzalez, A. cies. Syst. Biodiv., 16:497–511. & Martin-Gonzalez, A. 2001. Ciliate cryptobiosis: a microbial Reid, P. C. 1987. Mass encystment of a planktonic oligotrich cili- strategy against environmental starvation. Int. Microbiol., ate. Mar. Biol., 95:221–230. 4:151–157. Samarawickrema, N. A., Upcroft, J. A., Thammapalerd, N. & Gutierrez, J. C., Martın-Gonzalez, A. & Matsuoka, T. 1990. Upcroft, P. 2001. A rapid-cooling method for cryopreserving Towards a generalized model of encystment (cryptobiosis) in Entamoeba histolytica. Ann. Trop. Med. Parasitol., 95:853–855. ciliates: a review and a hypothesis. BioSystems, 24:17–24. Sheng, Y. L., He, M., Zhao, F. Q., Shao, C. & Miao, M. 2018. Phy- Hubalek, Z. 2003. Protectants used in the cryopreservation of logenetic relationship analyses of complicated class Spirotrichea microorganisms. Cryobiology, 46:205–229. based on transcriptomes from three diverse microbial eukary- Kim, S. M., Cho, J. B., Lee, E. H., Kwon, S. R., Kim, S. K., Nam, otes: Uroleptopsis citrina, Euplotes vannus and Protocruzia Y. K. & Kim, K. H. 2004. Pseudocohnilembus persalinus (Cilio- tuzeti. Mol. Phylogenet. Evol.. https://doi.org/10.1016/j.ympev. phora: Scuticociitida) is an additional species causing scuticocil- 2018.06.025. iatosis in olive flounder Paralichthys olivaceus. Dis. Aquat. Org., Simon, E. M. 1982. Breeding performance of Tetrahymena ther- 62:239–244. mophila following storage for 5 to 6 years in liquid nitrogen. Li, X. H., Yu, Y. H., Li, C. & Yan, Q. Y. 2018. Comparative study Cryobiology, 19:607–612. on the gut microbiotas of four economically important Asian Song, W. B. 2000. Morphological and taxonomical studies on carp species. Sci. China Life Sci., 61(6):86–95. some marine scuticociliates from China Sea, with description of Lovelock, J. E. 1953. The mechanism of the protective action of two new species, Philasterides armatalis sp. n. and Cyclidium glycerol against haemolysis by freezing and thawing. Biochim. varibonneti sp. n. (Protozoa: Ciliophora: Scuticociliatida). Acta Biophys. Acta, 11:28–36. Protozool., 39:295–322.

Song, W. B., Warren, A. & Hu, X. Z. 2009. Free-living Ciliates in the Zhao, X. L., Wang, Y. Y., Wang, Y. R., Liu, Y. F. & Gao, S. 2017. Bohai and Yellow Sea, China, 1st ed. Science Press, Beijing. Tetrahymena histone methyltransferase TXR1 is required for Tsutsumi, S., Watoh, T., Kumamoto, K., Kotsuki, H. & Matsuoka, both H3 and H3.3 lysine 27 methylation. Sci. China. Life Sci., T. 2004. Effects of porphyrins on encystment and excystment 60:264–270. in ciliated protozoan Colpoda sp. Jap. J. Protozool., 37:119–126. Zhao, Y., Yi, Z. Z., Warren, A. & Song, W. B. 2018. Species delim- Wang, C. D., Zhang, T. T., Wang, Y. R., Katz, L. A., Gao, F. & itation for the molecular taxonomy and ecology of a widely dis- Song, W. B. 2017a. Disentangling sources of variation in SSU tributed microbial eukaryotes genus Euplotes (Alveolata, rDNA sequences from single cell analyses of ciliates: impacts Ciliophora). Proceed. Roy. Soc. B., 285:20172159. of copy number variation and experimental errors. P. Roy. Soc. Zheng, W., Gao, F. & Warren, A. 2015. High-density cultivation of B-Biol. Sci., 284:20170425. the marine ciliate Uronema marinum (Ciliophora, Oligohy- Wang, Y. Y., Chen, X., Sheng, Y. L., Liu, Y. F. & Gao, S. 2017b. N6- menophorea) in axenic medium. Acta Protozool., 54:325–330. adenine DNA methylation is associated with the linker DNA of Zheng, W. B., Wang, C. D., Yan, Y., Gao, F., Doak, T. G. & Song, H2A.Z-containing well-positioned nucleosomes in Pol II-transcribed W. B. 2018. Insights into an extensively fragmented eukaryotic genes in Tetrahymena. Nucleic Acids Res., 45:11594–11606. genome: de novo genome sequencing of the multinuclear cili- Wilbert, N. 1975. Eine verbesserte Technik der Protargolimpragna-€ ate Uroleptopsis citrina. Genome Biol. Evol., 10:883–894. tion fur€ Ciliaten. Mikrokosmos, 64:171–179. Wolfe, J. & Bryant, G. 2001. Cellular cryobiology: thermodynamic and mechanical effects. Int. J. Refrig, 24:438–450. SUPPORTING INFORMATION Yan, Y., Xu, Y., Al-Farraj, S. A., Al-Rasheid, K. A. & Song, W. B. 2016. Morphology and phylogeny of three trachelocercids (Pro- Additional supporting information may be found online in tozoa, Ciliophora, Karyorelictea), with description of two new the Supporting Information section at the end of the article. species and insight into the evolution of the family Trachelocer- Table S1. One-way ANOVA of glycerol concentration cidae. Zool. J. Linn. Soc., 177:306–319. Zhang, T. T., Wang, C. D., Katz, L. A. & Gao, F. 2018. A paradox: effects. rapid evolution rates of germline-limited sequences are associ- Table S2. One-way ANOVA of thawing temperature ated with conserved patterns of rearrangements in cryptic spe- effects. cies of Chilodonella uncinata (Protist, Ciliophora). Sci. China Life Table S3. One-way ANOVA of the freezing duration Sci., 61:1071–1078. effects.