J. Insect Physiol. Vol. 43, No. 1, pp. 39–45, 1997 1997 Elsevier Science LtdPergamon Printed in Great Britain. All rights reserved0022-1910/97 $17.00 + 0.00PII: S0022-1910(96)00073-X

Ultrastructural Effects of Lethal Freezing onBrain, Muscle and Malpighian Tubules fromFreeze-Tolerant Larvae of the Gall Fly, EurostasolidaginisSTEPHEN D. COLLINS,* ALLAN L. ALLENSPACH,* RICHARD E. LEE JR.*†

Received 28 February 1996; revised 20 May 1996

In preparation for winter low temperatures, larvae of the gall fly, Eurosta solidaginis, accumu-late the cryoprotectants glycerol, sorbitol, and trehalose. The fat body cells of these freeze-tolerant larvae can survive intracellular freezing to −80°C for 48 h even though no wholelarvae survive this treatment. We hypothesized that some other tissue was more susceptibleto freezing and therefore may be responsible for larval death. This paper compares the ultras-tructure of brain, muscle, and Malpighian tubules between non-lethally frozen and lethallyfrozen freeze-tolerant larvae. The nuclei of cortical brain cells from lethally frozen larvaeexhibited clumped chromatin and nuclear membranes with occasional expansions or ‘blebs’of the intermembranous space, while the cytoplasm contained swollen spheres of endoplasmicreticulum. In contrast, non-lethally frozen brain contained nuclei with evenly dispersedchromatin, smooth nuclear membranes and a cytoplasm free of swollen endoplasmic reticu-lum. Muscle tissue of lethally frozen larvae contained disrupted myofilaments surroundingthe Z-line in comparison to non-lethally frozen muscle which had myofilaments extending allthe way to the Z-line. Alterations of Malpighian tubule cells from lethally frozen larvaeincluded an extracted cytoplasm with swollen and rounded mitochondria. In contrast, Mal-pighian tubule cells from non-lethally frozen larvae had a more concentrated cytoplasm withmany rod-shaped mitochondria. Results show alterations to all three tissue types due to lethalfreezing. The brain tissue contained the most observable alterations and therefore may be themost susceptible to lethal freeze damage. 1997 Elsevier Science Ltd. All rights reserved

Ultrastructure High pressure freeze fixation Insect Freeze-tolerant Comparative study

INTRODUCTION ice formation. In the autumn, the third instar larvaincreases its low temperature tolerance and becomes tol-During the summer, the female gall fly, Eurosta solidag-erant of extensive internal ice formation, in part, byinis (Diptera: Tephritidae), deposits her eggs within theaccumulating the cryoprotectants glycerol, sorbitol, andstems of the goldenrod plant, Solidago spp. Larvae over-trehalose (see Baust and Lee, 1981; Storey et al., 1981).winter in the third instar before they pupate and emerge

Although it was once believed, and is still generallyfrom the gall as an adult the following spring. Since thebelieved, that only extracellular freezing could be toler-gall usually remains exposed to all environmental con-ated by a freeze-tolerant organism, Salt (1959, 1962)ditions, the developing larvae must withstand the effectsreported that the fat body cells of E. solidaginis surviveof winter low temperatures.intracellular ice formation. A recent study by Lee et al.As a first, second, and early third instar, the larvae(1993) confirmed and extended Salt’s observations. Inhave not yet developed the capacity to survive internalthat study, fat body cells were frozen in vivo and in vitroin Grace’s medium with and without glycerol. It was dis-covered that fat body cells frozen in vivo (and sub-

*Department of Zoology, Miami University, Oxford, Ohio 45056,sequently removed and assessed for viability) and thoseU.S.A.frozen in vitro in 1m glycerol for 24 h at −80°C showed†Author for correspondence: Tel. 513-529-3141. Fax: 513-529-6900.

e-mail: lee@msmail.muohio.edu substantial survival even though ice formed intracellu-

39

40 STEPHEN D. COLLINSet al.

larly. Although larvae survive freezing between approxi- lobes of the brain, the muscle which operates the oralhook, and parts of the proximal segment of the Malpigh-mately −10 and −50°C, no whole larvae survived to com-

plete development to the adult stage after freezing at ian tubules were removed and fixed according to the fix-ation protocol listed below.−80°C. This suggests that damage to another cell type

(possibly nervous, muscular, and/or osmoregulatoryFreezing protocoltissues) may be the ‘weak link’ that is responsible for

larval death. Five larvae were removed from galls, observed forsigns of life (i.e., peristaltic movement of larva or move-Electron microscopy has been used in previous studies

to identify the effects of anoxia and/or freezing on tissue ment of the oral hook), and placed in a test tube fittedinto a larger test tube which was stoppered with a spongeultrastructure. For example, Singhal et al. (1988)

reported changes in nuclear and cytoplasmic components (to prevent condensation) and placed in a refrigeratedbath. Lethally frozen larvae were allowed to cool atof leech neurons due to anoxic or hypoxic treatment,

Sherman (1972) observed broken cristae and fractured 0.7°C/min to −55°C and held at that temperature for 48h. After freezing, larvae were placed on ice until dissec-membranes in mitochondria from frozen mice kidney,

and Fujikawa and Miura (1986) used electron tion. All larvae had been dissected within 15 to 25 minafter removal from the refrigerated bath.microscopy to study alterations in frozen plant tissue.

More recently, Morason et al. (1994) used the rela-High pressure freezing and freeze substitutiontively new technique of high pressure freezing with(HPF/FS) protocolfreeze substitution to compare the ultrastructure of fat

body cells between the freeze-tolerant and freeze-suscep- To minimize tissue degradation, non-lethally frozenlarvae and lethally frozen larvae were dissected whiletible gall fly larvae, E. solidaginis. It was shown that fat

body cells from freeze-tolerant larvae were more con- immersed in a 15% solution of dextran (see McDonaldand Morphew, 1993) in cold (4°C) Grace’s insectducive to cryofixation as compared to fat body cells from

freeze-susceptible larvae. Morason et al. (1994) attri- medium. Tissues were removed, isolated and placed intoplanchettes (Dahl and Staehelin, 1989) which had pre-buted this difference to the greater abundance of natural

cryoprotectants within the freeze-tolerant larvae. Other viously been coated with a 0.1% solution of soybean leci-thin in chloroform; lecitin makes it easier to remove thestudies (see Gilkey and Staehelin, 1986; Allenspach,

1993; Kiss et al., 1990; McDonald and Morphew, 1993) specimens from the planchettes after freeze substitution.Removal of tissue took approximately 3 to 7 min. Place-have shown that HPF/FS is better than chemical fixation

for preservation of cytoplasmic components and mem- ment of the tissue sample within the planchette wasfacilitated by immersing the planchette within the dis-branes. Using HPF/FS with electron microscopy, this

study compares the ultrastructural changes that occurred secting medium containing the sample and carefullymoving the sample above the planchette until the cohes-within brain, muscle, and Malpighian tubules of lethally

frozen (−55°C), freeze-tolerant larvae to the ultrastruc- ive force of the solution pulled the sample away fromthe forceps’ tip, allowing the sample to sink slowly intoture of non-lethally frozen (−22°C), freeze-tolerant larvae

of the gall fly. The objective of this study was to deter- the planchette. To improve cryofixation, air bubbles wereremoved from the tissue and planchette surface bymine which tissue is most susceptible to freeze-induced

alterations and therefore the possible ‘weak link’ manipulation with the tip of a fine dissecting pin.The large size of the brain relative to the planchettedetermining the lower limit of freeze tolerance in this

species. size required that it be subdivided into lobes (one lobeper planchette). Likewise, muscle tissue was cut into seg-ments in order to fit within the planchette, whereas 3 toMATERIALS AND METHODS6 mm segments of Malpighian tubule were folded to fit

Collection, storage, and preparation of larvae easily into a planchette. Generally, each planchette waspacked tightly to increase the effectiveness of cryofix-Freeze-tolerant larvae were collected on 16 January,

1993 at the Ecological Research Center, Miami Univer- ation.Cryofixation (i.e., HPF/FS) of samples was done insity, Oxford, Ohio. The larvae, still in their galls, were

placed in a −22°C freezer (a non-lethal freeze) until a Balzers HPM 010 high pressure freezing machine asdescribed by Moor (1987). Immediately after the freezingneeded. Prior to experimentation, the larvae were

removed from their galls, allowed to thaw, and observed event, the specimens (still in the planchettes) were placedin liquid nitrogen (LN2) until freeze substituted.for signs of life as judged by crawling or movement of

the oral hook. Non-lethally frozen larvae, frozen at High pressure frozen samples were freeze substitutedfor 3 days at −78°C in acetone containing 1.0% osmium−22°C, underwent no other manipulation prior to dissec-

tion and fixation of tissues. Lethally frozen larvae were tetroxide and 0.1% uranyl acetate in a customized alumi-num chamber (Allenspach, 1993). The solution washeld for 48 h at −55°C prior to dissection and fixation.

This freezing protocol (see below) was found to be the replaced daily. At the end of 3 days, the aluminumchamber was removed from the freezer and allowed tohighest subzero temperature which prevented all larvae

(n=25) from reaching the late pupal stage. The two optic passively warm to room temperature (which takes

41ULTRASTRUCTURAL EFFECTS OF FREEZING

approximately 165 min; see Allenspach, 1993). Thesamples were removed from the planchettes, rinsed indry acetone for 5 min, infiltrated and embedded in Epon-Araldite resin.

Microscopy

Thin sections were obtained with a diamond knifeattached to a Reichert Ultracut E microtome. Sectionswere stained for 5 min each with uranyl acetate followedby lead citrate. Viewing and photography were done ona JEOL 100S transmission electron microscope.

RESULTS

Cryofixation using high pressure freeze/freeze substi-tution (HPF/FS) yielded good preservation of all three FIGURE 2. Representative micrograph of portion of two neurons fromtissue types. Optimal cryofixation was observed in the the brain of larva that was frozen to −55°C for 48 h (a lethal freeze).

Note the clumped nuclear chromatin. The cytoplasmic matrix exhibitsperipherally located areas of all three tissue types. Withinsome precipitation and is not uniformly distributed. Swollen endoplas-these optimally cryofixed tissues, there were distinctmic reticula (arrowheads) are visible within the cytoplasm. N, nucleus.

ultrastructural differences in nuclear and/or cytoplasmic Bar =1 mmcomponents of lethally frozen cells as compared to thenon-lethally frozen cells. In lethally frozen larvae, cyto-plasmic differences involving the rough endoplasmicreticulum of brain cells and the accumulation of densebodies and swollen mitochondria in Malpighian tubulecells were noted while alterations in myofilament organi-zation were observed in muscle samples.

Nuclei of cortical brain cells occupied a large portionof the cell body when viewed in cross section (Figs 1 and2). Nuclear alterations of lethally frozen larvae includedclumping of the chromatin (Fig. 2), and irregular inter-membranous expansions, or ‘blebs’, of the nuclear envel-ope. Brain cell nuclei from non-lethally frozen larvaeexhibited evenly dispersed chromatin and less evidenceof membrane expansions (Figs 1 and 3).

Cytoplasm of non-lethally frozen brain cells was uni-formly granular and contained long, continuous profiles

FIGURE 3. A micrograph of a portion of the nucleus and cytoplasmof a neuron from the brain of a non-lethally frozen larva showing thelamellar appearance of the rough endoplasmic reticulum (arrowheads).Note also the evenly dispersed cytoplasmic ground substance. N,

nucleus. Bar = 0.5 mm

of rough endoplasmic reticulum as well as denselystained mitochondria (Fig. 3). Cytoplasm of lethallyfrozen brain was also uniformly granular with denselyosmiophilic mitochondria, but in contrast to non-lethallyfrozen brain, contained dilated endoplasmic reticulumdecorated with ribosomes (Fig. 4). Also present withinlethally frozen brain cytoplasm were darkly stained auto-phagic bodies (not pictured).

Muscle from non-lethally frozen larvae containedFIGURE 1. Representative micrograph of neuron from the brain of a myofilaments extending to the Z-line (Figs 5 and 6). Thenon-lethally frozen larva (i.e., frozen at −22°C). Note the evenly dis- most noticeable change within muscle tissue of lethallypersed nuclear chromatin. The cytoplasmic matrix is finely granular

frozen larvae compared to non-lethally frozen larvae waswith ribosomes dispersed throughout, a characteristic which reflectsthe appearance of an osmiophilic band bilaterallyoptimally cryopreserved tissue. Sparse profiles of endoplasmic reticu-

lum are evident (arrowheads). N, nucleus. Bar =1 mm apposed to the Z-line (Fig. 7) in some samples. Higher

42 STEPHEN D. COLLINSet al.

FIGURE 4. Higher magnification of a portion of the nucleus and cyto- FIGURE 7. Section through muscle of a lethally frozen larva. Noteplasm of a neuron from the brain of a lethally frozen larva. The the finely granular electron dense zones (arrowheads) and electron-chromatin is characteristically clumped in the nucleus and the nuclear translucent zones on each side of an irregular Z-line. mf, myofilaments.membrane is characterized by ‘blebs’ (asterisks). Note the swollen Bar = 0.5 mm.rough endoplasmic reticulum (arrowheads). N, nucleus. Bar = 0.5 mm.

magnification revealed that the area immediately adjacentto the Z-line to be granular and devoid of most myofila-ments (Fig. 8). The electron dense, osmiophilic bandassociated with the Z-line of lethally frozen larval muscleis not evident in non-lethally frozen samples (Fig. 5).

Differences between non-lethally frozen and lethallyfrozen Malpighian tubule cells were evident in both thenucleus and cytoplasm (Figs 9 and 10). Nuclear chroma-tin of the non-lethally frozen larvae was uniformly dis-persed (Fig. 9). In contrast, nuclear chromatin wasclumped in lethally frozen Malpighian tubule cells withmany of the aggregates positioned near the periphery ofthe nucleus (Fig. 10). Some aggregates were closely asso-ciated with darkly stained bead-like material. Unlike non-lethally frozen Malpighian tubule samples, which lackedswollen mitochondria, lethally frozen Malpighian tubulesamples contained mitochondria which were consistentlyFIGURE 5. Representative micrograph of a longitudinal sectionswollen and spherical; rod-shaped mitochondria found inthrough non-lethally frozen muscle fiber showing darkly stained Z-line

(Z) with adjoining myofilaments (mf). Bar = 0.5 mm. the non-lethally frozen samples were not evident in leth-

FIGURE 8. Higher magnification micrograph of Fig. 7 reveals areaFIGURE 6. Higher magnification of Fig. 5 identifying individual adjacent to Z-line devoid of most myofilaments (asterisks). Rarely aremyofilaments (arrowheads) extending to the Z-line. Bar = 0.25 mm. myofilaments associated with the Z-line. Bar = 0.25 mm.

43ULTRASTRUCTURAL EFFECTS OF FREEZING

malian tissue in which intracellular freezing typicallyresults in cell death (Mazur, 1984). This experiment isthe first to examine ultrastructural alterations due to alethal freeze within a naturally freeze-tolerant organism,the late third instar of the gall fly larvae, Eurosta solidag-inis, in which the fat body cells naturally survive intra-cellular freezing (Salt, 1959, 1962; Lee et al., 1993). Thecryofixation method used in the present study precludesalterations induced by chemical fixation and thus ensuresa more accurate picture of cell structure for these tissues.

Inherent with tissue freezing and thawing are othermechanisms with which the cell must successfully copein order to survive. Steponkus ((1984) p. 548) stated thatduring the freeze-thaw cycle “cells are subjected to amultitude of stresses including thermal, mechanical, andpossibly electrical perturbations.” Anoxia and/or hypoxiaare also expected to occur within the tissues since oxygenFIGURE 9. Micrograph of portion of nucleus (N) and surrounding

cytoplasm of Malpighian tubule cell from a non-lethally frozen larva. diffusivity is slow or non-existent through ice as com-Notice the diffuse nuclear chromatin. The cytoplasmic matrix is uni- pared to its diffusion through water. These numerousformly granular with no evidence of precipitation. Also note the pres- events naturally have led to various hypotheses as to the

ence of darkly stained, rod-shaped mitochondria (m). Bar =1 mm.primary cause of freezing injury. Cellular dehydration (aresult of water leaving the cell to join extracellular ice),and the ensuing changes in solute concentration and cellvolume, has often been implicated as the primary causeof injury (Mazur, 1984). Mechanical mechanisms suchas sharp-edged ice fronts which pierce the cell membraneor walls of ice that squeeze cells between them have alsobeen suggested (Steponkus, 1984). During thawing,tissues must also withstand osmotic fluxes across cell andorganelle membranes as the water that was drawn out ofthe cell during the freeze re-enters the cell or organelleduring the thaw. It was the purpose of this study tocharacterize the cellular morphological alterations thatoccur as a result of lethal freezing and subsequent thaw-ing in the freeze-tolerant gall fly larvae. These results arecompared to previous studies of freeze-induced alter-ations as well as to studies of hypoxia/anoxia-induced

FIGURE 10. Micrograph of portion of a nucleus and surrounding alterations since frozen tissues are also subjected tocytoplasm of Malpighian tubule cell from a lethally frozen larva. The

hypoxia/anoxia during freezing.electron dense bodies in the nucleus represent clumped chromatin.All three tissue types showed differences between theMitochondria (m) appear swollen and electron translucent, and numer-

ous dense bodies (arrowheads) are evident in the cytoplasm. N, non-lethally frozen and the lethally frozen samples. Anucleus. Bar =1 mm. noticeable alteration to lethally frozen muscle involved

disintegration of many of the myofilaments adjoining theZ-line. This band of amorphous, electron dense materialally frozen samples. Densely osmiophilic spheresappears to be evidence of denatured myofilaments. A(possibly autophagic vesicles) were apparent in both non-similar phenomenon was reported to occur in postmortemlethally frozen and lethally frozen Malpighian tubuletuna muscle (Davie and Sparksman, 1986); that musclesamples, but appeared to be more abundant in the cyto-had been examined after time delays in which the muscleplasm of the lethally frozen samples (Fig. 10).had likely been exposed to anoxia. The tissue had notbeen re-frozen as in experiments described here. Davie

DISCUSSIONand Sparksman (1986) suggested that the denaturationwas the result of a lower pH brought on by increasedPrevious studies have reported changes in nuclear

chromatin, the nuclear membrane, and cytoplasmic anaerobic glycolysis. Since the intracellular pH of E. sol-idaginis larva increases with decreasing temperaturecomponents such as mitochondria and endoplasmic retic-

ulum as a result of either freezing or anoxia (for (Storey et al., 1984, 1986), this change may play a rolein the morphological alterations we observed.examples see Singhal et al., 1988; Sherman, 1971, 1972;

Sherman and Liu, 1982; and Fujikawa and Miura, 1986). The lethally frozen Malpighian tubule cells showedclumps of nuclear chromatin with many of the clumpsMost of these studies, however, have been with mam-

44 STEPHEN D. COLLINSet al.

associated with a string of bead-like material. The stress with death resulting only when the stress is toosevere. Of course, subtle alterations or ones not observ-chromatin clumping is similar to that seen in lethally

frozen brain nuclei but the consistent association of the able with the electron microscope may be responsible forcell death.clumps with bead-like strings in the Malpighian tubule

samples makes it uncertain whether these clumps are the In this experiment, larvae were frozen to a subzerotemperature that no larvae could survive. To answer theresult of lethal freezing or are natural structures related

to polytene chromosomes. Sorsa (1984) observed similar question of which changes are reversible, in futureexperiments larvae could be frozen to the highest subzerobead-like structures believed to be polytene chromo-

somes in the salivary gland of Drosophila. More reliable temperature at which alterations first begin to appear.Larvae could then be allowed to rewarm and continueevidence of injury are the many swollen mitochondria

with fragmented cristae and the greater abundance of development with samples removed at intervals andobserved with the electron microscope. This would allowdense bodies within the lethally frozen samples of Mal-

pighian tubules. Sherman (1972) noted swollen mito- the experimenter to see which alterations occur first andwhich alterations are reversible.chondria with fragmented cristae as a result of freezing

in mice kidneys. Singhal et al. (1988) found fragmented Due to the complex environment of the cell and themany unobservable processes occurring within it, it ismitochondrial cristae in leech neurons exposed to anoxia.

Disrupted cristae suggest impaired mitochondrial func- difficult by observation alone to state with certaintywhich tissue is suffering the most damage due to a lethaltion (Sherman, 1971). Accumulation of dense bodies,

judged to be autophagic vacuoles, may also be indicative freeze. Electron microscopy, however, proved effectivein showing the salient changes to brain, muscle, and Mal-of cell injury.

The ultrastructural alterations within the larval brain pighian tubule cells in response to a lethal freeze and istherefore a good starting point for future studies.exposed to lethal freezing were similar to ones reported

by Wegener (1987) in his study of adult fly brain exposedto anoxia. Most noticeable in this study, and in common REFERENCESwith Wegener’s study, were the many swellings of the

Allenspach A.L. (1993) Ultrastructure of early chick embryo tissuesrough endoplasmic reticulum, expansions of the nuclear after high pressure freezing and freeze substitution. Microsc. Res.membrane, and clumping of the nuclear chromatin. It Tech. 24, 369–384.

Baust J.G. and Lee R.E. (1981) Divergent mechanisms of the frost-seems that anoxia and freezing produce many similarhardiness in two populations of the gall fly, Eurosta solidaginis. J.ultrastructural alterations; what is less certain is whetherInsect Physiol. 27, 485–490.these observable alterations could lead to cell and ulti-

Baust J.G. and Nishino M. (1991) Freezing tolerance in the goldenrodmately larval death. If it is assumed that any alterations gall fly (Eurosta solidaginis). In: Insects at Low Temperature (Edshinder cell survival, the brain, having the most noticeable Lee R. E. and Denlinger D. L.), pp. 260–275. Chapman and Hall,

New York.alterations, would appear to be the ‘weak link’ of the lar-Dahl R. and Staehelin L.A. (1989) High-pressure freezing for the pres-vae.

ervation of biological structure: theory and practice. J. ElectronClumping of nuclear chromatin suggests that transcrip-Microsc. Tech. 13, 165–174.

tion and translation would be adversely affected since the Davie P.S. and Sparksman R.I. (1986) Burnt tuna: an ultrastructuralorderly arrangement of the chromosomes has been alt- study in muscle of the yellowfin tuna (Thunnus albacares) caught

on rod and reel and southern bluefin tuna (Thunnus maccoyii)ered. Singhal et al. (1988) argued that stacked endoplas-caught on handline or longline. J. Food Sci. 51, 1122–1128.mic reticulum, not swollen and rounded endoplasmic

Fujikawa S. and Miura K. (1986) Plasma membrane changes causedreticulum, is necessary for protein synthesis. In the studyby mechanical stress in the formation of extracellular ice as a pri-

of adult fly brain exposed to anoxia, Wegener (1987) mary cause of slow freezing injury in fruit-bodies of basidio-noted that the changes brought on by anoxia are revers- mycetes (Lyophyllum ulmarium (Fr.) Kuhner). Cryobiology 23,

371–382.ible if the treatment is not too severe. Kalinina and Gro-Gilkey J.C. and Staehelin L.A. (1986) Advances in ultrarapid freezingmov (1991) discovered that the membrane-damaged

for the preservation of cellular ultrastructure. J. Electron Microsc.nucleus from a frozen amoeba was capable of directingTech. 3, 177–210.

cellular activity when placed in a non-frozen, enucleated Herman M.M., Miquel J. and Johnson M. (1971) Insect brain as aamoeba of the same species, although the ability of the model for the study of aging — Age related changes in Drosophila

melanogaster. Acta neuropath. (Berl.) 19, 167–183.nucleus to direct reproduction had been lost. As reportedKalinina L.V. and Gromov D.B. (1991) The nucleus of non-viableby Lee et al. (1993), large changes of lipid droplet struc-

frozen-thawed ameoba has normal morphology and high functionalture within the fat body cells from frozen Eurosta larvaeactivity (nuclear transplantation and electron microscopical

are not indicative of cell death. These observations reveal studies). Cryo-letters 12, 87–92.that cell death doesn’t automatically result even when Kiss J.Z., Giddings Th.H., Staehelin L.A. and Sack F.D. (1990) Com-

parison of the ultrastructure of conventionally fixed and high press-major morphological alterations occur. Some alterationsure frozen/freeze substituted roots tips of Nicotiana and Arabi-may be reversible while others may be a normal partdopsis. Protoplasma 157, 64–74.of development (see Morason et al., 1994 regarding the

Lee R.E., McGrath J.J., Morason R.T. and Taddeo R.M. (1993) Sur-accumulation of autophagic vacuoles in fat body cells of vival of intracellular freezing, lipid coalescence and osmotic fra-developing Eurosta larvae) or cell maintenance. Alter- gility in fat body cells of the freeze-tolerant gall fly Eurosta solida-

ginis. J. Insect Physiol. 39, 445–450.ations, therefore, may be a normal cellular response to a

45ULTRASTRUCTURAL EFFECTS OF FREEZING

Mazur P. (1984) Freezing of living cells: mechanisms and impli- (Erpobdellidae): ultrastructural studies. J. Invert. Pathol. 52,409–418.cations. Am. J. Physiol. 247, C125–C142.

Sohal R.S. and Sharma S.P. (1972) Age-related changes in the fineMcDonald K. and Morphew M.K. (1993) Improved preservation ofstructure and number of neurons in the brain of the housefly, Muscaultrastructure in difficult-to-fix organisms by high pressure freezingdomestica. Exp. Geront. 7, 243–249.and freeze substitution: 1. Drosophila melanogaster and Strongylo-

Sorsa V. (1984) Electron microscopic mapping and ultrastructure ofcentrotus purpuratus embryos. Micros. Res. Tech. 24, 465–473.Drosophila polytene chromosomes. In: Insect ultrastructure (EdsMoor H. (1987) Theory and practice of high pressure freezing. In:King R. C. and Akai H.), pp. 75–110. Plenum Press, New York.Cryotechniques in Biological Electron Microscopy (Eds Stein-

Steponkus P.L. (1984) Role of the plasma membrane in freezing injurybrecht R. A. and Zierold K.), pp. 175–191. Springer. Berlin.and cold acclimation. Annu. Rev. Plant Physiol. 35, 543–584.Morason R.T., Allenspach A.L. and Lee R.E. (1994) Comparative

Storey K.B., Baust J.G. and Storey J.M. (1981) Intermediary metab-ultrastructure of fat body cells of freeze-susceptible and freeze-tol-olism during low temperature acclimation in the overwintering gallerant Eurosta solidaginis larvae after chemical fixation and highfly larva, Eurosta solidaginis. Comp. Physiol. 144, 183–190.pressure freezing. J. Insect Physiol. 40, 155–164.

Storey K.B., McDonald D.G. and Booth C.E. (1986) Effect of tempera-Rees D. and Usherwood P.N.R. (1972) Effects of denervation on theture acclimation on haemolymph composition in the freeze-tolerantultrastructure of insect muscle. J. Cell Sci. 10, 667–682.larvae of Eurosta solidaginis. J. Insect Physiol. 32, 897–902.Salt R.W. (1959) Survival of frozen fat body cells in an insect. Nature

Storey K.B., Miceli M., Butler K.W., Smith I.C.P. and Deslautiers R.184, 1426.(1984) 31P-NMR studies of the freze-tolerant larvae of the gal fly,Salt R.W. (1962) Intracellular freezing in insects. Nature 193, 1207–Eurosta solidaginis. Eur. J. Biochem. 142, 591–595.1208.

Wegener G. (1987) Insect brain metabolism under normoxic andSherman J.K. (1971) Correlation in ultrastructural cryoinjury of mito-hypoxic conditions. In: Arthropod Brain: its Evolution, Develop-chondria with aspects of their respiratory function. Exp. Cell Res.ment, Structure and Functions (Ed. Gupta A. P.), pp. 369–397.66, 378–384.John Wiley and Sons.Sherman J.K. (1972) Comparison of in vitro and in situ ultrastructural

cryoinjury and cryoprotection in mitochondria. Cryobiology 9,112–122.

Sherman J.K. and Liu K.C. (1982) Ultrastructure before freezing, while Acknowledgements—Support was provided by the NSF-IBNfrozen, and after thawing in assessing cryoinjury of mouse epididy- #9305809 (R.E.L.), by NSF grants DIR-8820387 (A.L.A.) and BSR-mal spermatozoa. Cryobiology 19, 503–510. 861424 to the Electron Microscope Facility. This study partially ful-

Singhal R.M., Sarnat H.B. and Davies R.W. (1988) Effects of anoxia filled requirements for the Master of Science Degree to S. D. Collins.We appreciate the helpful comments of two anonymous reviewers.and hyperoxia on the neurons in the leech Nephelopsis obscura