Seasonal changes in the low-temperature behaviour of xylem ray parenchyma cells in red osier dogwood (Cornus sericea L.) with respect to extracellular freezing and supercooling

Micron,Vol. 27, No. 3--4,pp. 181-191, 1996

Copyright © 1996 Elsevier Science Ltd Printed in Great Britain. All rights remrved 0968-4328/96 $15.00 + 0.00

Pergamon PII: S0968-4328(96)00031-5

Seasonal Changes in the Low-temperature Behaviour of Xylem Ray Parenchyma Cells in Red Osier Dogwood (Cornus sericea L.) with Respect to Extracellular Freezing and Supercooling SEIZO FUJIKAWA, .1 KATSUSHI K U R O D A t and JUN OHTANVf *Institute of Low Temperature Science, Hokkaido University, Sapporo 060, Japan t Faculty of Agriculture, Hokkaido University, Sapporo 060, Japan (Received18 January 1996; accepted21 May 1996)

Al~tract--Tbe low temperature behaviour of xylem ray parenchyma cells in red osier dogwood (ComussericeaL.) was examined by differential thermal analysis (DTA), eryo-seanning electron microscopy, freeze-fracture replica electron microscopy and a survival assay (leakage of electrolytes). DTA provided a profile that is typical of extracellular freezing in xylem ray parenehyma ceils in both winter and summer. Observations of xylem ray parenchyma ceils by electron microscopy indicated, however, that lowtemperature behaviour was different from that predicted by DTA. Electron microscopy revealed that, upon cooling at 0.1 °C/rain, the ray parenchyma ceils in winter exhibited typical extraceilular freezing, whereas cells in summer exhibited intracellular freezing below - 15*C. Cooling at 1.25°C/day ( < 0.001*C/rain) produced a slight collapse of the cell walls as a result of partial dehydration, but it did not inhibit the intraceilular freezing in xylem ray parenchyma cells in summer. It is suggested that failure of DTA to reveal the low temperature exotherm (LTE) upon intracellular freezing was due to an overlap of temperature between the high temperature exotherm (HTE) and the LTE, in addition to a reduction in the LTE by the partial dehydration of cells. It is concluded that red osier dogwood has xylem ray parenchyma cells whose low-temperature behaviour changes from extracellular freezing in winter to supercooling in snmmer, possibly as a result of seasonal differences in permeability of the cell walls to water. This type of seasonal change in the low-temperature behaviour may produce a superior mechanism for the adaptation to freezing temperatures of cells of plants growing in cold regions, in which dehydration tolerance also changes seasonally. Copyright © 1996 Elsevier Science Ltd.

Key words: Comussericea,freezing tolerance, supercooling, xylem ray parenchyma cell, cryo-scanning electron microscopy, freezefracture replica.

INTRODUCTION Plant tissues respond in a poikilothermic fashion to changes in environmental temperature. With a reduction in temperature to slightly below 0°C, ice forms outside the living cells and expands in apoplastic spaces. Cells in many plant species, from temperate and cold climates, exploit mechanisms that allow them to survive freezing temperatures. Such mechanisms involve either freezing avoidance or freezing tolerance (Burke et al., 1976; George et al., 1982). Plant cells that survive low temperatures by freezing avoidance display characteristics of deep supercooling. During freezing, the cellular water in such cells remains liquid to near the homogeneous nucleation temperature of water (-40°C) by remaining isolated from extracellular ice. Below -40°C, supercooling is disrupted with resultant intracellular freezing and consequent injury. Plant cells that survive low temperatures by the development of freezing tolerance lose their cellular water to apoplastic spaces in response to extracellular ice. As the temperature of the tissue declines, such cells become progressively dehydrated and extracellular 1 Corresponding author. Fax: 81-11-706-7142. E-mail: sfuji@lt. hines.hokudai.ac.jp

freezing occurs (Levitt, 1980). Some plant cells are known to be able to survive even at the temperature of liquid nitrogen by extracellular freezing that circumvents lethal intracellular freezing (Sakai, 1960). Xylem ray parenchyma cells of woody species have been reported to adapt to low temperatures either by deep supercooling or by extracellular freezing, depending upon the species (George et al., 1982; Sakai and Larchar, 1987). Comprehensive groupings of woody species, based on differences in the low-temperature behaviour of xylem ray parenchyma cells, have been reported. There is a distinct trend in the distribution of woody species to cold regions that depends on differences in the low-temperature behaviour of the xylem ray parenchyma cells (George et al., 1974, 1982; Becwar and Burke, 1982). Woody species that have xylem ray parenchyma cells that exhibit deep supercooling are restricted to regions where the annual minimum temperature does not fall below -40°C, whereas woody species that have xylem ray parenchyma cells that exhibit extracellular freezing are found in colder regions. In order to characterize the low-temperature behaviour of xylem ray parenchyma cells, differential thermal analysis (DTA) has been used extensively in previous studies (Quamme et al., 1973; George et al., 1974; Kaku and Iwaya, 1978). In DTA, a high temperature exotherm 181

s. Fujikawa et al.

182

(HTE) is first produced at around - 5 ° C , being due to freezing of water in apoplastic spaces. The HTE gradually disappears with a gradual reduction in temperature. The xylem of some woody species does not produce a further exotherm at lower temperatures, whereas xylem of others yields a new exotherm upon a further reduction in temperature. The new exotherm produced at lower temperatures is called the low temperature exotherm (LTE) and it is postulated that the LTE is produced by intracellular freezing of xylem ray parenchyma cells upon the breakdown of deep supercooling. In DTA, extracellular freezing or deep supercooling is defined by the presence or absence of the LTE (Quamme et al., 1973; George et al., 1982). However, the LTE is often unclear, rendering a distinction between supercooling and extracellular freezing difficult. Furthermore, the LTE often has two or more peaks for reasons that remain unclear (Rajashekar and Burke, 1978; Kaku and Iwaya, 1978; Rajashekar et al., 1982). Low-temperature techniques for electron microscopy, such as cryo-scanning and freeze-fracture electron microscopy, have the potential to reveal directly the low-temperature behaviour of xylem ray parenchyma cells under freezing conditions without confusion by artifacts. Using these techniques, we showed previously that, in &flax obassia, the LTE corresponds clearly to the occurrence of intracellular freezing upon the breakdown of supercooling in the xylem ray parenchyma cells. We also provided unequivocal results showing that each ray cell acts as an isolated unit upon freezing, that supercooling ability is not affected by the presence or absence of water in the adjacent vessel lumen, and that it is the cell wail, and not plasma membrane, that is responsible for the supercooling ability (Fujikawa et al., 1994). Red osier dogwood was identified as a tree with xylem ray parenchyma cells that exhibit extracellular freezing in a previous study using DTA (George et al., 1974). However, recent electron microscopic studies suggested the possibility that the xylem ray parenchyma cells in red osier dogwood might have unexpected low-temperature behaviour unlike that indicated by DTA (Malone and Ashworth, 1991; Ristic and Ashworth, 1994). The present electron microscopic study by cryo-scanning and freeze-fracture electron microscopy was designed to obtain direct visual evidence of the low-temperature behaviour of the xylem ray parenchyma cells of red osier dogwood under freezing conditions. Our results indicate that the low-temperature behaviour in xylem ray parenchyma cells of red osier dogwood differs not only from that predicted by DTA but also from that deduced from previous studies by electron microscopy.

Hokkaido University, in summer (July to August) and in winter (December to February), from 1993 to 1995.

MATERIALS AND METHODS

Assay of freezing injury

Mater&l Three-year-old twigs were collected from red osier dogwood (Comus sericea), grown on the campus of

DTA Fresh debarked twig internodes were cut into lengths of 3 cm and utilized as fresh xylem. Oven-dried xylem, after heating at 120°C overnight, was used as a reference. These samples of xylem were split longitudinally into two parts near the center. The junction of a 36-gauge copper-constantan thermocouple was placed in contact with one of the halves of the split xylem, sealed with the other split half and the entire sample was wrapped in alminium foil. A pair of one fresh and one oven-dried sample was placed in a glass bottle, the bottle was closed and placed in a freezer (MDF-192; Sanyo Co. Ltd, Tokyo) that was equipped with a programmable controller (ES-100P; Tajiri Co. Ltd, Sapporo). For DTA, samples were equilibrated at 3°C for 30 min and cooled to - 5 0 ° C at rates of 0.1, 0.01 and 0.005°C/ min. The temperature of freezing events was determined from the difference between the output from the fresh sample and that from the oven-dried reference sample with a hybrid recorder (HR-1300; Yokogawa Co. Ltd, Tokyo).

Preparation of samples for electron microscope Fresh xylem was cut into small blocks (about 3 x 3 × 4 mm in the radial, tangential and longitudinal directions, respectively). A small block was fixed to a freeze-fracture holder with starch paste, in an arrangement that revealed the tangential plane of the spring wood regions that had developed in the previous year (relative to the time of sampling) upon fracturing for electron microscopy. These samples were placed in petri dishes. They were equilibrated at 3°C in a programmable freezer and then they were cooled to desired temperatures at a cooling rate of 0.1 °C/rain. For cooling at 1.25°C/day (< 0.001°C/min), samples were first kept at - 5 ° C for one day after seeding with ice and then cooled in a stepwise manner, at steps of 1.25°C/day, to the desired temperature. Preliminary examination showed that the stepwise cooling brought about the same results as linear cooling at a rate of 0.001 °C/min that was obtained by using the programmable controller. These samples were cryofixed by abrupt immersion in melting freon 22 ( - 150°C) soon after the desired temperature had been reached (samples cooled at 0.1 °C/ min) or after equilibration for 1 day at the desired temperature (samples cooled at 1.25°C/day). As controis, fresh samples that had been equilibrated at room temperature (22°C) were cryofixed.

Occurrence of freezing injury was assessed by measuring the leakage of electrolyte from damaged cells. To measure such leakage, fresh xylem was cut into small blocks (1 × 1 × 4 mm). From 8 to 12 blocks were

Low-temperatureBehaviourof XylemRay ParenchymaCells placed at the bottom of a test tube. The test tube was equilibrated at - 5 ° C , seeded with ice at - 5 ° C and cooled at a rate of 0.1°C/rain in the programmable freezer or frozen in a stepwise manner at 1.25°C/day to the desired temperature, as in the case of samples for electron microscopy. Frozen samples in test tubes were thawed at 0°C for 4-12 hr soon after the desired temperature had been reached (samples cooled at 0.1 °C/min) or after equilibration for 1 day at the desired temperature (samples cooled at 1.25°C/day). Two millilitres of distilled water were added to each test tube which was then shaken overnight at 4°C. The electroconductivity of each solution was measured at 22°C with a conductivity meter (CD-35MII; M and S Instruments Inc., Tokyo). The material was then boiled for 10 min and shaken overnight and then electroconductivity was determined again. The electroconductivity before boiling was converted to the percentage of that after boiling to give the relative electronconductivity (Fujikawa and Miura, 1986).

Electron microscopy Cryofixed samples were processed for observation with a cryo-scanning electron microscope (Cryo-SEM, JSM 840A equipped with CRU-40; JEOL Co. Ltd, Tokyo) by the method described previously by Fujikawa (1991). In brief, a cryofixed sample was transferred to a cold stage in the specimen preparation chamber of the Cryo-SEM, equilibrated, fractured, etched for 10 s, and evaporated with platinum-carbon and then with carbon at - 105°C. For deep-etching, fracture faces were etched for 3 min. After observations and photographic recording on a cold stage of Cryo-SEM at -164°C, an evaporated film (freeze-replica) was extracted from the sample for preparation of freeze-replicas. Freezereplicas were also produced independently with a freeze-etching apparatus (JFD 7000; JEOL Co. Ltd). These replicas were examined with a transmission electron microscope (1200EX; JEOL Co. Ltd). For revealing the ratio of cellular shrinkage, the longaxis to short-axis ratio of the fractured cytoplasm in the ray parenchyma cells was calculated from the cryo-SEM photographs. The number indicated is the mean+SD that was measured from 40 cells in each condition.

Light microscopy Thin sections (about 1 #m in thickness) were cut with glass knives on an ultramicrotome (Ultracut Om U4; Reichert, Vinna, Austria) from samples that had been embedded in epoxy resin after chemical fixation with 4% glutaraldehyde and I% OsO4 at 4°C for thinsection electron microscopy (Fujikawa and Takabe, 1996). After staining with 1% safranin, the sections were examined with a light microscope (BHS; Olympus Co. Ltd, Tokyo).

183

RESULTS

DTA The low-temperature behaviour of xylem ray parenchyma cells in red osier dogwood was examined by DTA during lowering of the temperature at 0.1 °C/min (Fig. 1). DTA produced an HTE at around - 5 ° C . The HTE gradually decreased with lower temperatures and disappeared completely at around - 2 2 to -25°C. Neither any shoulder in the HTE nor any new exotherm corresponding to an LTE was detected with a further decrease in temperature to -50°C. No distinct differences were detected in the DTA profiles between xylem harvested in winter (Fig. la) and in summer (Fig. lb). Similar DTA profiles were also obtained by cooling at 0.01°C/min (Fig. lc) and 0.005°C/min (Fig. ld) in both winter (not shown) and summer, although the temperatures for occurrence and disappearance of the HTE gradually shifted to warmer temperatures at the lower cooling rates. These results of DTA suggested, as in the previous studies, that xylem ray parenchyma cells of red osier dogwood responded to low temperatures by extracellular freezing throughout all the seasons and at all the cooling rates tested.

Ultrastructure of ray parenchyma cells Xylem ray tissue of red osier dogwood, as viewed in tangential sections in the springwood region developed in the past year (relative to the time of sampling), consisted of a single row of a small number of elongated cells at the upper and lower margins and of a double row of more circular, ellipsoidal cells in the center (Fig. 2a). Observations by electron microscopy were also made of similar areas that were revealed by freeze-fracturing. No distinction was made during observations of low temperature behaviour among the locations of cells within a ray tissue because there were no differences in location. The long-axis to short-axis ratio of ray parenchyma cells located in the center of ray tissue was 2.6-t-1.3 in the control that was cryofixed from room temperature. Cryofixation of xylem from room temperature resulted in conversion of the cellular water into very small intracellular ice crystals, which yielded well preserved cytoplasmic structures (Fig. 2b-d). Intracellular ice crystals produced by cryofixation were slightly larger in summer (about 50 nm in mean diameter) than in winter (about 20 nm in mean diameter) samples. In winter, Cryo-SEM (Fig. 2b) could not detect existence of small ice crystals in the cytoplasm, but freeze-replica (Fig. 2d) revealed formation of small ice crystals as a result of cryofixation by showing granular fracture faces in the cytoplasm that were produced by sublimation of small ice crystals. In summer, Cryo-SEM showed the presence of small ice crystals as a result of cryofixaton. The granular appearance that was produced by the sublimation of small ice crystals was especially

184

S. Fujikawa et al.

seen in vacuoles (Fig. 2c). The cytoplasm in winter samples contained numerous small organelles (Fig. 2b), while that in summer samples contained comparatively few large organelles (Fig. 2c). The plasma membranes of ray parenchyma cells both in winter and in summer showed the even distribution of intramembrane particles (IMPs) throughout entire fracture faces (Fig. 2d). No distinct differences in the ultrastructure of plasma membranes between winter and summer samples were observed. Freeze-fracture replicas showed that the endoplasmic reticulum (ER) in winter was vesicular or reticular in form (Fig. 2d), while that in summer was composed mainly of reticular or sheet-like cisternae (not shown).

Low-temperature behaviour of ray parenchyma cells in winter Cooling of xylem in winter at 0. l°C/min resulted in the gradual shrinkage of ray parenchyma cells (Fig. 3), producing finally typical extracellular freezing around all the cells (see Fig. 5). Upon cooling to - 10°C, where DTA had already revealed the start of freezing of water in the apoplastic spaces (Fig. 1), cells showed only slight shrinkage (long-axis to short-axis ratio, 5.0-I-2.4) due to partial dehydration (Fig. 3a). Cryofixation converted cellular water into very small intracellular ice crystals. The presence of ice crystals was suggested o~ly by the distinct granularity of the cytoplasm which was a result of sublimation of ice by etching (Fig. 3b, compare with Fig. 3d). Freezing to - 3 0 ° C produced more distinct shrinkage (long-axis to short-axis ratio, 13.1+6.8) of the ray parenchyma cells (Fig. 3c). Cryoflxation did not produce detectable intracellular ice crystals in these cells (Fig. 3d). The cytoplasm yielded smooth fracture faces after etching, indicating that all freezable water had been lost by dehydration. The shrinkage of the cytoplasm was accompanied by distinct collapse of the cell walls (Fig. 3a and Fig. 3c). Slower cooling at 0.005°C/min (not shown) produced the same results as cooling at 0.1°C/min. Thus, cooling at a rate of 0.1°C/min corresponded to equilibrium freezing in the case of winter samples. The plasma membrane in winter had the same structure as those in controls, that were cryofixed from room temperature, even after freezing to - 3 0 ° C (Fig. 3d). However, freezing of winter samples produced dramatic changes in the organization of the ER, After freezing to -30°C, vesicular and reticular ER were lost from the cytoplasm. Instead, extensive sheet-like cisternae of ER in single or multiple layers were produced beneath the plasma membranes (Fig. 3d). The changes in the organization of the ER were already evident after freezing to - 1 0 ° C at 0. l°C/min, revealing development of sheet-like cisternae of ER and cylindrical ER (Fig. 3b), which might represent a process of rearrangement of the ER.

Low-temperature behaviour of ray parenchyma cells in summer Cooling of xylem in summer at 0.1 °C/min resulted in low-temperature behaviour that was completely different from that in winter (Fig. 4). Cooling to - 1 0 ° C revealed that cells had very similar outlines to those in controls. Cryofixation converted cellular water into small intracellular ice crystals similar to those in controls (not shown, but similar to Fig. 2c). Cooling to -20°C, by contrast, produced large intracellular ice crystals that ranged from 1 #m to more than 10/zm in diameter within a slightly shrunken cytoplasm (Fig. 4a), although DTA did not detect these changes (Fig. I). The cytoplasmic materials were confined among large intracellular ice crystals and identification of individual organelles was difficult. It seems reasonable that the very large ice crystals were produced by intracellular freezing during cooling at a rate of 0.1°C/rain, and not by cryofixation. In order to determine the cooling rate that might induce extraceUular freezing in the ray parenchyma cells in summer, we examined the effects of very low cooling rates. Cooling at 1.25°C/day (< 0.001°C/rain) produced shrinkage of cells at -7.5°C, without formation of large intracellular ice crystals due to intracellular freezing (Fig. 4b). Individual cytoplasmic organelles were clearly identified as a result of the absence of large intracellular ice crystals. Deep etching removes ice crystals by sublimation and produces large holes in the cytoplasm

a

-8

-22

b -7.5

4-a

©

-25

-7

-6

-15

-12.5

I

I

I

0

-10

-20

I

-30

I

-40

Freezing temperature (°C) Fig. 1. DTA profiles during cooling of xylem. The numbers next to each profile indicate the temperatures of initiation (left) and termination (right) of the HTE. All the profiles are means from three preparations. (a) Winter sample cooled at 0. l °C/min. (b) Summer sample cooled at 0.1°C/rain. (c) Summer sample cooled at 0.01°C/rain. (d) Summer sample cooled at 0.005°C/ min.

Low-temperatureBehaviourof XylemRay ParenchymaCells corresponding to the intracellular ice crystals, if they occurred. However, deep-etching did not produce such a structural change in the cytoplasm, indicating the absence of large intracellular ice crystals in ray parenchyma cells that were cooled at 1.25°C/day to -7.5°C (not shown). Additionally, deep-etching did not produce empty spaces between the cell wall and shrunken cytoplasm, indicating the absence of plasmolysis. Freezing below - 1 0 ° C started to produce large intracellular ice crystals (more than 0.6/zm in diameter) within a slightly shrunken cytoplasm (Fig. 4c and 4d), demonstrating the occurrence of intracellular freezing. Upon freezing below -30°C, all cells froze intracellularly (Fig. 5). These phenomena were unchanged even after prolonged incubation of summer samples at -7.5°C for 4 days during cooling at 1.25°C/day to below - 10°C. Freeze-fracture replicas showed that distinct ultrastructural changes in plasma membranes were produced by freezing below - 5 ° C , in the case of cooling at 1.25°C/day, and by freezing below -7.5°C, in the case of cooling at 0.1°C/min. At both cooling rates, ultrastructural changes in the plasma membranes were observed before intracellular freezing occurred (Fig, 5). The characteristics of the ultrastructural changes in the plasma membranes, produced by slow freezing, were common to both cooling rates. The plasma membrane produced aparticulate domains, namely, areas free of or with only rare IMPs (Fig. 4e). The aparticulate domains in the plasma membranes were frequently accompanied by fracture-jumps, namely, areas where the fracture plane deviated locally at sites of closely approached intracellular membranes (Fig. 4e). Survival

Survival of ray parenchyma cells, as determined from leakage of electrolytes, is shown in Fig. 6. In winter, freezing at 0.1°C/min resulted in high survival even at -30°C. In summer, however, survival was abruptly reduced by freezing, with distinct differences between cooling at 0.1°C/rain and at 1.25°C/day. Cooling at 1.25°C/day resulted in a reduction in survival at higher temperatures than cooling at 0.1°C/min, with ELs0 (temperature for 50% electrolyte leakage) at - 4 ° C and -8°C, respectively. At both cooling rates, almost complete destruction (100% electrolyte leakage) occurred at temperatures above those at which intracellular freezing occurred, showing that a reduction in survival in summer was not related to the initiation of intracellular freezing (compare Fig. 5 with Fig. 6). Reduction of survival was rather related to the occurrence of ultrastructural changes in the plasma membranes (Fig. 4e).

DISCUSSION Xylem ray parenchyma cells in red osier dogwood have been reported to respond to low temperatures by

185

extracellular freezing (George et al., 1974). No seasonal changes in the low-temperature behaviour of xylem ray parenchyma cells have been reported in many woody species that exhibit extracellular freezing (George et al., 1982). The results of our DTA also support previous results, showing that xylem ray parenchyma cells in red osier dogwood, both in winter and summer, yield typical DTA profiles that appear to represent extracellular freezing (Fig. 1). Our examination by cryo-scanning and freeze-fracture electron microscopy showed, however, that the low-temperature behaviour of xylem ray parenchyma cells was different from that predicted by DTA. Electron microscopy revealed that xylem ray parenchyma cells in winter exhibited typical extracellular freezing (Fig. 3) as predicted by DTA, whereas those in summer exhibited intracellular freezing (Fig. 4). Thus, the low temperature behaviour of xylem ray parenchyma cells of red osier dogwood seems to change with the season. It is suggested that the electron-microscopic techniques that we used do not introduce artifacts into the structures of ray parenchyma sells in the frozen state, with the exception that cellular water in the'liquid state is converted into small ice crystals as a result of cryofixation. Because the cryofixation technique that we employed converted the cellular water of xylem ray parenchyma cells into very small ice crystals, it was not hard to distinguish between the very small ice crystals ( < 50 nm in diameter) produced as a result of cryofixation (Fig. 2) and the large ice crystals (>0.6 #m) produced as a result of intracellular freezing during slow freezing (Fig. 4). It is suggested that the failure of DTA to detect the exotherm (LTE) that reflects intracellular freezing of ray parenchyma cells in red osier dogwood in summer may be due to an overlap in temperatures between the LTE and the HTE, in addition to the reduction of the LTE due to partial dehydration of cells. With cooling at 0.1°C/min, the HTE continued to around - 2 5 ° C (Fig. 1) while intracellular freezing, as observed by electron microscopy, occurred between - 15 and - 2 0 ° C (Fig. 5). A similar overlap was also noted when a lower cooling rate was used (compare Fig. 1 with Fig. 5). Electron microscopy can reveal the true low-temperature behaviour of the xylem ray parenchyma cells without effect of freezing in apoplastic spaces. Malone and Ashworth (1991) and Ristic and Ashworth (1994) studied the low-temperature behaviour of xylem ray parenchyma cells in red osier dogwood using a modified freeze-substitution electron-microscopic technique, in which freeze-substituted specimens were rehydrated at room temperature before embedding in resin to facilitate penetration of resin into cells (Ristic and Ashworth, 1993). Their studies revealed that the cell walls of the ray parenchyma cells did not collapse upon slow freezing, while the protoplasm was distinctly shrunken, with consequent plasmolysis. They suggested that such plasmolysis did not produce injury in winter, whereas plasmolysis was accompanied by injury due to tearing of the cytoplasm in summer. These results are

186

S. Fuiikawa et al.

Fig. 2. Control structures of xylem ray parenchyma cells. Samples were cryofixed with equilibration to room temperature, except for (a) which was chemically fixed at 4°C. (a) Light micrograph showing ray tissue, x 350. (b) Cryo-SEM photograph showing cells in winter, x 4000. (c) Cryo-SEM photograph showing cells in summer, x 5000. (d) Freeze-replica photograph showing part of a cell in winter. PF, Protoplasmic fracture face in the plasma membrane. ER, Endoplasmic reticulum. V, Vacuole, x 83,500.

Low-temperature Behaviour of Xylem Ray Parenchyma Cells

187

I Fig. 3. Structure of xylem ray parenchyma cells in winter after cooling at 0.1 °C/rain to the indicated temperature and cryo-fixation. (a) Cryo-SEM photograph showing cells that had been cooled to - 10°C, × 4200. (b) Freeze-replica photograph showing part of a cell that had been cooled to - 10°C. × 48,400. (c) Cryo-SEM photograph showing cells that had been cooled to - 30°C, × 3200. (d) Freeze-replica photograph showing part of a cell that had been cooled to -30°C. See Fig. 2 for abbreviations, × 49,100.

188

S. Fujikawa e t al.

N

Fig. 4. Structure of xylem ray parenchyma cells in summer after slow cooling to the indicated temperature and cryo-fixation. (a) Cryo-SEM photograph showing cells that had been cooled at 0.1°C/rain to -20°C. An asterisk indicates some intracallular ice crystals, x 6500. (b) Cryo-SEM photograph showing cells that had been cooled at 1.25°C]day to - 7.5°C, x 7000. (c) Freeze-replica photograph showing cells that had been cooled at 1.25°C/day to -30°C. For asterisks, see Fig. 4a, x 5000. (d) Freeze-replica photograph showing cells that had been cooled at 1.25°C/day to -20°C, EF, Exoplasmic fracture face in the plasma membrane. CW, C_.eUwall. For definition of PF, see Fig. 2 and asterisks, see Fig. 4a, x 21,100. (e) Freeze-replica photograph showing part of cell that had been cooled at 1.25°C/day to - 5°C. For definition of PF, see Fig. 2. Arrows show some fracture jumps, x 66,000.

Low-temperature Behaviour of Xylem Ray Parenchyma Cells

completely different from our present results. We observed the distinct collapse of cell walls in winter (Fig. 3), as well as the slight collapse of cell walls even in intracellularly frozen cells in summer (Fig. 4). The method for preparation of samples used by Ristic and Ashworth (1993) might generate artifacts that allow return of the collapsed cell wails to the state before freezing by rehydration, with consequent artificial plasmolysis, It is also suggested that tearing of the cytoplasm in summer might be due to combination of artifactual recovery of the cell walls and the occurrence of intracellular freezing. Our results obtained by cryoscanning and freeze--fracture electron microscopy may not be confused by such artifacts because the observed structures had been kept in a frozen state in the presence of ice. We conclude that, in winter, the xylem ray parenchyma cells were extracellularly frozen, with distinct collapse of cell walls, although it remains unclear how collapse of the rigid cell walls occurs. Deep supercooling characteristic of the xylem ray parenchyma cells of woody species has been defined by DTA, when the LTE was produced by cooling generally at rates between 1.0 and 0.1°C/min (George et al., 1974; Kaku and Iwaya, 1978). Gusta et al. (1983) indicated, however, that when these tissues were exposed to subzero temperatures for prolonged periods, the LTE moved to lower temperatures (e.g. in river bank grape and scarlet oak) or completely disappeared (e.g. in red ash). On the other hand, exposure of the xylem in S t y l a x obassia to - 2 0 ° C for 4 days did not bring any changes of the LTE that appeared around -30°C (our unpublished results). These results indicate that deep supercooling characteristic of the ray parenchyma cells covers a wide variety of ranges, possibly reflecting differences in permeability to water of ray parenehyma

~

loo

~

80

N ©

60

~

40

~

20

U

....t

J,,

X

4J

~

(

t ~

0 0

~ -10

-20

-30

Freezing temperature (%) Fig. 5. Frequency of intracellularly frozen cells as a function of cooling rate and freezing temperature. ( S ) Winter cells cooled at 0.1°C/min. (O) Summer cells cooled at 0.1°C/rain. ( x ) Summer cells cooled at 1.25°C /day. Each symbol is derived from observations of one sample by Cryo-SEM where more than 100 cells were observed.

,/•.•X

100

Of-

X

o

80

go

4o

N

2o

/"

0

-10

-20

-30

Freezing temperature (°C) Fig. 6. Survival of cells as a function of cooling rate and freezing temperature. (Q) Winter cells cooled at 0.1 °C/min. (O) Summer cells cooled at 0. l°C]min. ( x ) Summer cells cooled at 1.25°C/day. Sampling was performed on the same day for both sets of summer cells (O, x). Each symbol refers to one preparation.

cells upon exposure to subzero temperatures (Ashworth and Abeles, 1984; Fujikawa et al., 1994). From these results, we concluded that xylem ray parenchyma cells of red osier dogwood in summer could be included in the category of deep supercooling tissue because they were supercooled to around - 2 0 ° C at 0.1 °C/rain. Moreover, cooling at 1.25°C/day (<0.001°C/min), even with additional exposure to -7.5°C for 4 days, did not inhibit the intracellular freezing that occurred below 10°C (Fig. 5). In summer, the limit of the supercooling to around -20°C is a general feature of xylem ray parenchyma cells in many woody species that are categorized as exhibiting deep super cooling (Wisniewski and Ashworth, 1986; Arora et al., 1992). Supercooling of xylem ray parenchyma cells has been considered to be a mechanism for prevention of freezing injury (Hong et al., 1980; Ashworth et al., 1983; Montane et al., 1987; Arora et al., 1992), while small and large discrepancies between temperature for the occurrence of injury and an LTE have also been reported (Kaku and Iwaya, 1978; Lindstrom et al., 1995). Lindstrom et al. (1995) suggested factors that might limit survival, other than the occurrence of intracellular freezing, upon the breakdown of deep supercooling. In the case of red osier dogwood in summer, the ray parenchyma cells were injured before initiation of intracellular freezing (Figs 1 and 5). The reduction in survival might be due to the partial dehydration of cells during freezing. Partial dehydration of xylem ray parenchyma cells in red osier dogwood produced aparticulate domains with accompanying fracture-jump lesions in the plasma membranes (Fig. 4). Similar structural changes were produced by freezing-induced dehydration in cortical parenchyma -

0

189

190

S. Fujikawa et al.

cells of mulberry (Fujikawa, 1994), in protoplasts from rye and oat leaves (Webb et al., 1994), in cells of rye leaves (Webb and Steponkus, 1993) and in leaves of Arabidopsis thaliana (Uemura et al., 1995). It is suggested that these structural changes in plasma membranes were produced by close apposition of membranes due to freezing-induced dehydration (Fujikawa and Miura, 1986; Steponkus and Lynch, 1989; Fujikawa, 1994) and led to injury (Fujikawa, 1995). Our present results show that freezing injury was more or less reduced in xylem ray parenchyma cells of red osier dogwood in summer that were cooled at 0.1 °C/ min, as compared those cooled at <0.001°C/min (Fig. 6), suggesting a protective role for supercooling, to some extent, via a reduction in the extent of dehydration. The response of xylem ray parenchyma cells in summer to freezing-induced partial dehydration contrasted with the case in winter, when neither reduction in survival (Fig. 6) nor ultrastructural changes in the plasma membrane (Fig. 3) were produced by even more severe dehydration due to extracellular freezing. The different response of the cells to freezing-induced dehydration in winter might reflect the acquisition of freezing tolerance, a result of cold acclamation that involves many complex physiological and metabolic changes in the cells (Levitt, 1980; Sakai and Larchar, 1987). Distinct changes in chemical, physical and structural properties of the plasma membrane during cold acclamation have been reported in close association with the development of freezing tolerance (Yoshida, 1984). Augmentation of cytoplasmic organelles in xylem ray parenchyma ceils in winter (Fig. 2) is also associated with the process of cold acclimation in many woody species (Pomeroy and Siminovich, 1971; Niki and Sakai, 1981; Fujikawa and Takabe, 1996). Conversion of vesicular ER to the sheet-like cisternae that form multiplex lamellae upon exposure to freezing to near subzero temperatures, as also observed in this study (Fig. 3), has been postulated to be one of the mechanisms for adaptation to dehydration in cortical parenchyma cells of mulberry after cold acclimation (Fujikawa and Takabe, 1996). It is suggested that the difference in low-temperature behaviour, namely, extracellular freezing or supercooling, in xylem ray parenchyma cells of red osier dogwood between winter and summer xylem may be due to changes in permeability to water of the cell walls as a result of cold acclimation. The difference between the capacity for deep supercooling and the capacity for extracellular freezing of xylem ray parenchyma cells among different woody species has been considered to be largely a reflection of differences in permeability of the cell walls to water (Quamme et aL, 1973; Ashworth et al., 1983; Fujikawa et al., 1994). Wisniewski et al. (1987) suggested that key sites for the permeability to water of the cell walls of xylem ray parenchyma cells might be the pit regions through the vessels. It is also suggested that enzymatic alterations in structures in the pit areas, including protective (amorphous) layers, largely affect the appearance of the LTE (Wisniewski

and Davis, 1989; Wisniewski et a/., 1991). Further studies will be necessary to identify the critical features of the cell walls in xylem ray parenchyma cells of red osier dogwood that lead to the different low-temperature behaviours of winter and summer samples, which might possibly reflect the cold acclimation of winter samples. In conclusion, it is suggested that seasonal changes in the low-temperature behaviour, from supercooling in summer to extracellular freezing in winter, might provide a superior mechanism for adaptation to freezing temperatures of cells of plants growing in cold regions. Supercooling in summer can prevent or minimize cellular dehydration when cells have little tolerance to dehydration, while extracellular freezing in winter can allow cells to adapt to lower temperatures when cells have a high degree of tolerance to dehydration. We need to understand the effects of cold acclimation in the plant cells, considering more fully the association between changes in permeability of cell walls to water and changes in cellular tolerance to freezing-induced dehydration. Acknowledgement--K. K. is a research fellow of the Japan Society for the Promotion of Science.

REFERENCES Arora, R., Wisniewski, M. E. and Scorza, R., 1992. Cold acclamation in genetically related (sibling) deciduous and evergreen peach (Prunus persica [L.] Batsch), I. Seasonal changes in cold hardiness and polypeptides of bark and xylem tissues. Plant Physiol., 99, 1562-1568. Ashworth, E. N., Rowse, D. J. and Billmyer, L. A., 1983. The freezing of water in woody tissues of apricot and peach and the relationship to freezing injury. J. Am. Soc. Hortic. Sci., 108, 299-303. Ashworth, E. N. and Abeles, F. B., 1984. Freezing behavior of water in small pores and the possible role in the freezing of plant tissues. Plant Physiol., 76, 201-204. Becwar, M. R. and Burke, M. J., 1982. Winter hardiness limitations and physiography of woody timberline flora. In: Plant Cold Hardiness and Freezing Stress, Li, P. H. and Sakai, A. (eds), Academic Press, New York, pp. 307-323. Burke, M. J., Gusta, L. V., Quamme, H. A., Weiser, C. J. and Li, P. H., 1976. Freezing and injury in plants. Ann. Rev. Plant Physiol., 27, 50%528. Fujikawa, S., 1991. Freeze fracture technique. In: Electron Microscopy in Biology: A Practical Approach, Harris, J. R. (ed.), IRL Press, Oxford, pp. 173-201. Fujikawa, S., 1994. Seasonal ultrastructural alterations in the plasma membrane produced by slow freezing in cortical tissues of mulberry (Morus bombycis Koids. cv Goroji). Trees, 8, 288-296. Fujikawa, S., 1995. A freeze-fracture study designed to clarify mechanisms of freezing injury due to the freezing-induced close apposition of membranes in cortical parenchyma cells of mulberry. Cryobiology, 32, gA.A ~54. Fujikawa, S. and Miura, K., 1986. Plasma membrane ultrastractural changes caused by mechanical stress in the formation of cxtracellular ice as a primary cause of slow freezing injury in fruit-bodies of basidiomycctes (Lyophyllum ulmarium Fr. kithner). Cryobiology, 23, 371-382. Fujikawa, S., Kuroda, K. and Fukazawa, K., 1994. U1trastructural study of deep supercooling of xylem ray parenchyma cells from Stylax obassia. Micron, 25, 241-252. Fujikawa, S. and Takabe, K., 1996. Formation of multiplex lamellae by equilibrium slow freezing of cortical parenchyma cells of mulberry and its possible relationship to freezing tolerance. Protoplasma, 190, 189-203. George, M. F., Becwar, M. R. and Burke, M. J., 1982. Freezing avoidance by deep undercooling of tissue water in winter-hardy plants. Cryobiology, 19, 628-639.

Low-temperature Behaviour of Xylem Ray Parenchyma Cells George, M. F., Burke, M. J., Pellet, H. M. and Johnson, A. G., 1974. Low temperature exotherms and woody plant distribution. Hort. Sci., 9, 519-522. Gusta, L. V., Tyler, N. J. and Chen, TH.-H., 1983. Deep undercooling in woody taxa growing north of the - 4 0 ° C isotherm. Plant Physiol., 72, 122-128. Hong, S., Sucoff, E. and Lee-Stadelmann, O., 1980. Effect of freezing of deep supercooled water on the viability of ray cells. Bot. Gaz., 141, 464--468. Kaku, S. and Iwaya, M., 1978. Low temperature exotherms in xylems of evergreen and deciduous broad-leaved trees in Japan with reference to freezing resistance and distribution range. In: Plant Cold Hardiness and Freezing Stress, Li, P. H. and Sakal, A. (eds.), Academic Press, New York, pp. 227-239. Levitt, J., 1980. Responses of Plant to Environmental Stresses, vol. 1, Academic Press, New York. Lindstrom, O. M., Anisko, T. and Dirr, M. A., 1995. Low-temperature exotherms and coold hardiness in three taxa of deciduous trees. J. Amer, Soc. Hort. Sci., 120, 830-834. Malone, S. R. and Ashworth, E. N., 1991. Freezing stress response in woody tissues observed using low-temperature scanning electron microscopy and freeze substitution techniques. Plant Physiol., 95, 871-881. Montano, M., Rebhuhn, M., Hummer, K. and Lagerstedt, H. B., 1987. Differential thermal analysis for large-scale evaluation of pea cold hardiness. Hort. Sci., 22, 1335-1336. Niki, T. and Sakai, A., 1981. Ultrastractural changes related to frost hardiness in the cortical parenchyma cells from mulberry twigs. Plant Cell Physiol., 22, 171-183. Pomeroy, M. K. and Siminovich, D., 1971. Seasonal cytological changes in secondary phloem parenchyma cells in Robinia pseudoacacia in relation to cold hardiness. Can. J. Bot., 49, 787795. Quamme, H., Weiser, C. J. and Stushnoff, C., 1973. The mechanism of freezing injury in xylem of winter apple twigs. Plant physiol., 51, 273-277. Rajashekar, C. and Burke, M. J., 1978. The occurrence of deep undercooling in the genera Pyrus, Prunus, and Rosa: A preliminary report. In: Plant Cold Hardiness and Freezing Stress, Li, P. H and Sakai, A. (eds), Academic Press, New York, pp. 213-225.

191

Rajashekar, C., Westwood, M. N. and Burke, M. J., 1982. Deep supercooling and cold hardiness in genus Pyrus. Amer. Soc. Hort. Sci., 107, 968-972. Ristic, Z. and Ashworth, E. N., 1993. New infiltration method permits use of freeze substitution for preparation of woody tissues for transmission electron microscopy. J. Microsc., 171, 137-142. Ristic, Z. and Ashworth, E. N., 1994. Response of xylem ray parenchyma cells of red osier dogwood (Comus sericea L.) to freezing stress: Microscopic evidence of protoplasm contruction. Plant Physiol., 104, 737-746. Sakai, A., 1960. Survival of the twig of woody plants at -196°C. Nature, 185, 393-394. Sakai, A. and Larchar, W., 1987. Frost Survival of Plants, Responses and Adaptation to Freezing Stress. Springer-Verlag, Berlin. Steponkus, P. L. and Lynch, D. V., 1989. Freeze/thaw-induced destabilization of the plasma membrane and the effects of cold aclimation. J. Bionerg. Biomemb., 21, 21-41. Uemura, M., Joseph, R. A. and Steponkus, P. L., 1995. Cold acclimaton of Arabidopsis thaliana. Plant Physiol., 109, 15-30. Webb, M. S. and Steponkus, P. L., 1993. Freeze-induced membrane ultrastructural alterations in rye (Secale cereale) leaves. Plant Physiol., 101, 955-963. Webb, M. S., Uemura, M. and Steponkus, P. L., 1994. A comparison of freezing injury in oat and rye: Two cereals at the extremes of freezing tolerance. Plant Physiol., 104, 467-478. Wisniewski, M. and Ashworth, E. N., 1986. Seasonal variation in deep supercooling and dehydrative resistance. Hort. Sci., 21,503-505. Wisniewski, M., Ashworth, E. N. and Schaffer, K., 1987. The use of lanthanum to characterize cell wall permeablity in relation to deep supercooling and extracellular freezing in woody plants: II. Intergenetic comparisons between Betula lenta and Betula papyrifera. Protoplasma, 141, 160-168. Wisniewski, M. and Davis, G., 1989. Evidence for the involvement of a specific cell wall layer in regulation of deep supercooling of xylem parenchyma. Plant Physiol., 91, 151-156. Wisniewski, M., Davis, G. and Schaffer, K., 1991. Mediation of deep supercooling of peach and dogwood by enzymatic modifications in cell-wall structure. Planta, 184, 254-260. Yoshida, S., 1984. Chemical and biophysical changes in the plasma membrane during cold acclamation of mulberry bark ceils (Morns Bombycis Koids, cv Goroji). Plant Physiol., 76, 257-265.