- Email: [email protected]
Differential scanning calorimetric analysis of antifreeze protein activity in the common mealworm, Tenebrio molitor
Biochimica et Biophysica Acta, 957 (1988) 217-221 Elsevier
217
BBA33250
Differential scanning calorimetric analys|s of antifreeze protein activity in the common mealworm, Tenebrio molitor T h o m a s N. H a n s e n and John G. Boast Center of Cryobiological Research, State University of New York, Binghamton, N Y (U.S.A.)
(Received 20 January 1988) (Revised manuscript received 16 May 1988)
Key words: Antifreeze protein: Protein activity: DSC, differential scanning calorimetry; (7". mofitor)
Antifreeze proteins (AFP) are able to inhibit the growth of ice-crystals at temperatures below the equilibrium freezing poin,~ (Tf) of hemolymph. The analysis of AFP activity has commonly involved the use of direct microscopic observation of a sample following inoculation with ice. The resulting activity, defined as the amount of thermal hysteresis observed between Tf and the subsequent rapid growth of ice, has been reported to range up to 7 o C. However, most studies report high level of variation, possibly due to ice-crystal size variability and the presence of non-visible ice nuclei. We describe a new method of analysis of AFP activity using differential scanning calorimetry (DSC). DSC analysis reveals much higher activity., up to 10°C, with less variation observed within a sample, and is not subject to the difficulty of accurate assessment of ice-crystal volume.
introduction Antifreeze proteins (AFP) form unique classes of macromolecules able to inhibit the growth of ice crystals. These proteins have beea isolated in a variety of species, including fish [1], spiders [2] and insects [3]. The fish AFP have been most extensively studied, and at least four different classes of the protein have been reported based on the presence or absence of carbohydrate sidegroups and on amino acid content [46]. All of the antifreeze molecules lower the hysteretic freezing point (Thy), a secondary onset of crystalliza-
Abbreviations: AFP, antifreeze proteias; Tf, equilibrium freezing point; Thy, hysteretic freezing point; Tm, equilibrium melting point; DSC, differential scanning calorimetry Correspondence: J.G. Baust, Center for Cryobiological Research; State University of New York, Binghamton, NY 13901, U.S.A.
tion in the presence of an ice nucleus. The Thy is lowered in a non-colligative manner: the melting point of the solution is changed in a predicable colligative manner while the T,y is lowered below the equilibrium melting point [7]. in a solution devoid of AFP, the equilibrium melting point (Tm ) and equilibrium freezing point (Tf) are equivalent; however, in an AFP-containing solution the Tf cannot be readily defined due to the action of the proteins. The proposed mechanism for AFP activity is based on the glycoprotein in AFP model. These proteins contain may repeating subunits of Ala-Ala-Thr with a disaccharide linked to the threnonmc. The overall structure forms an a-hello cal rod with a polar side-chain cvc~y 4~ A~ a distance which also separates the oxygen atoms in ice. The polar groups are thought to hydrogen bond with the oxygen atoms in the nascent icecrystal, thereby 'poisoning' the lattice [7]. The protein-crystal interaction would increase the ice front's angle of curvature and disrupt the free-en-
0167-4838/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
218
ergy relationships between bulk water and ice. Thus, a lower temperature for the continued growth of the ice front would be required. The first antifreeze activity was observed in the common mealworm, Tenebrio molitor [8]. Studies revealed the activity to be caused by a protein of molecular weight between 10000-12000 and to be inactivated by trypsin [9]. Further work revealed the possibility of multiple peptides and purified forms were found to not contain carbohydrate side-groups [10-12]. The most common methods of studying the AFP require direct microscopic observation of the sample as it is cooled either on a cold stage [13] or in a capillary immersed in an alcohol bath [14]. The fluid sample is inoculated with an ice crystal (either directly or by whole-sample freezing followed by a partial melt) and then slowly cooled until crystal growth is observed. In the absence of AFP, growth is observed almost immediately (~< 0.02* C) upon cooling below Tm. In the presence of the antifreeze, the growth is delayed as the temperature drops until the AFP activity is overcome and an almost explosive growth of long ice spicules is observed. Different scanning calorimetry (DSC) has been extensively used to study the effect of freezing and thawing on biological samples [15]. By utilizing power-compensated DSC which compares an inert reference to the sample, the enthalpic changes in the sample over temperature a n d / o r time can be monitored. Such changes include exothermic and endothermic events as well as glass transitions either during chemical reactions or during state changes. We propose a new method of analysis of AFP activity using DSC. We suggest that this mode of analysis lends a greater quantitative measure by allowing the ice-crystal volume to be directly measured and controlled. Materials and Methods Samples from four animal species were collected and analyzed for antifreeze activity. Hemolymph was collected from common mealworm larvae ( T. molitor)~ which were maintained at 25 ° C under a 16:8 light/dark photoperiod. Hernolymph from the gall fly third instar larvae, Eurosta solidaginis (October field collections from native
populations) was obtained by cuticular puncture. Plasma from adult female mice (C3H/MeMs x 129/J)F1 C31 (23°C, 14:10 L / D ) and from pigeons, Columba lioia (25°C, 16:8 L / D ) was also collected. Solutions of 0.9% NaCI and 20 mM Tris buffer (pH 7.0) containing 0.5 M NaCI were also analyzed. Samples of 0 . 2 / d or less were suspended in oil (Cargille Type B) and examined under a light microscope connected to a cold stage (Clifton Nanoliter Milliosometer, 100 × magnification). The samples were cooled to - 4 0 ° C, and slowly warmed (0.10 C/min). The temperature at which the last ice-crystal disappeared was noted. The samples were then recooled to - 4 0 0 C ano warmed to a temperature that maintained the smallest visible ice crystal (area = 25 #mZ). Samples were then slowly cooled (0.1 o C/min) and the temperature at which ice-crystal growth resumed was noted. A difference between the hold temperature and that of ice-growth (Thy; 'onset' temperature) was taken as the functional definition of antifreeze activity (hold-onset - AFP activity). Additional samples between 1 and 5 #1 were placed in oil-filled aluminium pans for analysis by DSC. The pans were placed in a Perkin-Elmer DSC-7 equipped with an lntracooler II. An empty aluminum pan was placed in the reference cell. The samples were cooled and warmed at 1 ° C/min and the crystallization and melting points noted. Samples were then cooled to - 4 0 " C (10" C/min), held for 5 min, warmed to various partial melt temperatures between - 1 0 and O°C in 0.1*C increments, and held for 5 min at each increment to allow for ice-protein interaction and system stabiliz~tion. The samples were then slowly cooled to - 1 5 o C (1 ° C/min), and the onset and area of the freeze exotherm calculated. The onset was observed as the start of the cwstalhzation exotherm. The area was calculated from the onset point to the point where the heat content returned to pre-crystallization levels. The ice-inoculated freeze exotherm was compared to the melt endothermic area to calculate the percent ice present in the sample. Oil baselines were run before and after sample analysis. A Perkin-Elmer 7700 computer was used to calculate the areas under the curves and the onset temperature of the exotherms. Thermograms were normalized |'or sam-
219
pie weight. Values are presented as mean _+ standard deviation. H
Results
/:6
v
A summary of microscopic observations of the antifreeze activity are presented in Table I. Individual samples were tested multiple times for activity. A wide range of activity was noted in the AFP-containing samples but not in the samples devoid of AFP, even though careful attention to uniform ice-crystal size was attempted. The T. molitor samples were observed to behave in a manner similar to other AFP, showing an explosive growth of long spicules at Thy. In DSC experiments samples were layered between oil for antifreeze activity measurements. Thermograms of a typical 7'. molitor hemolymph sample are presented in Fig. 1. A delay in the onset of the freeze exotherm was observed to increase with the higher hold temperatures. The area under the curves also increased with the higher temperatures, indicating that less ice was present. In order to test for possible effects by other components of the hemolymph, analysis of AFP-free gall fly hemolymph and of plasma samples was undertaken. Thermograms obtained from E. solidaginis hemolymph are presented in Fig. 2. There were no delay in the onset of the freeze exotherm. Mouse plasma, 0.9% NaCI and Tris buffer showed shailar results to that of the gall fly, with no delay in :he freeze exotherm when ice was
/: _9.o
"-4 a
-1~4
I
-12
I
-10
-8
.~
I
-4
.
2
0
Ternperoture (C) Fig. 1. DSC thermograms of T. mofitor hemolymph illustrating AFP activity. The sample was frozr~,n and allowed to undergo a partial melt before refreezing. The three temperatures presented are - 2.3, - 2.2 and - 2.1 ° C. The - 2.10 C hold (point H) and exotherm onset (point O) are marked. The delay in freeze onset increased as the te~nperature increased. The sampie's crystallization temperatu~'e was - 23.3 5:1.5 ° C (n = 3).
present in the sample. Thermograms obtained from pigeon plasma are presented in Fig. 3. The thgrmograms were similar to the other samples devoid of antifreeze. The oil baselines showed no thermal activity. When a single T. molitor sample was run multiple times to the same hold temperature, some variation in the delay of the exotherm was observed ( - 1 . 2 1 _+ 0.17°C; range - 1 . 0 3 to - 1.50°C; n =
2,0
.?
~ 1.0
TABLE I MICROSCOPIC EXAMINATION OF SAMPLES F O R ANT I F R E E Z E ACTIVITY All samples were analyzed using a cold stage connected to a light microscope and were run multiple times upon inoculation with the smallest ice-crystal observable at 100x(25 #me).
"r
0.0
, -14
, -12
-b
.
~
Temperature
Sample ( . ) a
Hold-freeze_+ S.D. ( ° C )
Range ( * C )
T. molitor (8) T. mofitor (5)
0.70 + 0.21 0.98 + 0.29
0.40-1.01 0.29-1.27
Pigeon Plasma (3) Mouse Plasma (3) Tris Buffer (3)
0.02 + 0.01 0.02 + 0.01 0.01 + 0.01
0.01-0.02 0.01-0.02 091-0.02
" Number of times same sample was tested for antifreeze activity.
.~,
.~
-~
-
o
(C)
Fig. 2, DSC thermograms of E. solidaginis hemolymph illustrating the behavior of a solution devoid of AFP. The sample was frozen and partially melted ( - 2 . 3 , - 2 . 2 and - 2 . 1 ° C ) before being refrozen. The - 2 . 1 ° C hold and exotherm onset are marked by an arrow. The melt exothermic area increased as the temperature increased, indicating less ice was present. The freeze onset was immediate upon cooling in the presence of ice. The sample crystallization temperature was - 17.8+0.9 (n = 3).
220 3,0
12u 10
~8
.-. 2.0
f,
' 6
& O & a
•
O
o ,1
_~1.0
OC
.... ~
,,•
"~2
I
"10
I
I
"8 "6 Temperoture (C)
I
I
"4
"2
,,,
0
Fig. 3, DSC thermograms of pigeon plasma. The sample was frozen and partially melted ( - 3.0, - 2 . 0 and - 1 . 4 ° C) before being refrozen, The hold and onset exotherm for the - 1 . 4 ° C point are marked by an arrow (point S: shoulder on the exotherm). The sample crystallization temperature was - 1 5 + 0.33 = C (n -- 3).
6). The variability was however only a fraction of that obtained by microscopic analysis. The DSC data for T. molitor hemolymph are presented in Fig. 4. The amount of AFP activity and the amount of ice melt are plotted against the hold temperature. There was an increase in the amount of sample melted with increased temperature. There was a parallel increase in AFP activity also with increased temperature. Two hold temperatures ( - 0 . 9 and - 0 . 7 ° C) showed increased melt over adjacent points. The reason in unexplained, but the AFP activity followed the points. "U i
- 8o
10-
60
i
40 E #
20
o <
0
O_ -3
-2
-I
0
HOld (*C)
Fig. 4. T. molitor. Antifreeze activity and per cent sample melt versus actual hold temperature. The amount of sample that melted increased with higher temperature. AFP activity (calculated by subtracting the hold/annealing temperature from the onset of the freeze exotherm) also increased with higher temperature. Two points ( - 0 . 9 and - 0 . 7 ° C ) had higher melts than adjacent points, but also revealed higher activity.
% Ice
Fig. 5. Antifreeze activity versus per cent ice in the ~ample. Samples not containing any antifreeze activity (Tris buffer (~), 0.9% NaCI (<>), and Pigeon Plasma (n) had a difference between the hold and onset of no more than 0.1 ° C. Samples containing antifreeze activity (7". molitor: 0 , &) had a high difference between the hold temperature and the onset of the exotherm. The pigeon plasma difference increased at 15% ice due to a shoulder on the exotherm curve.
All the DSC data are summarized in Fig. 5. The calculated onset was subtracted from the hold temperature and the difference plotted against the percent ice present in the samples (as indicated by the area under the curves). None of the samples exhibited any AFP activity except for the 7". molitor hemolymph. The pigeon plasma showed a delay in freeze onset at 15% ice but not show any delay at lower ice content. An inspection of the thermograms revealed an inflection pnint midway through the freeze, wlfich the computer interpreted as the freeze onset (Fig. 3). Discussion
The microscopic examination for AFP activity revealed a large range of values for activity of repeated individual samples and between samples containing the protein, even when the ice-crystals were carefully monitored for 'constant size' (Table I). A range of activity values has been reported by others [1]. Samples devoid of AFP were, however, found to be very constant in ice-crystal growth characteristics. Analysis of AFP activity by DSC was conducted in a manner similar to the microscopic examinations. Samples containing AFP showed a delay in the freeze onset in the presence of ice (Fig. 1). The delay was observed to increase in an almost exponential manner. Similar results of
22|
increased activity with decreased crystal size have been reported [16], although ice-crystal fraction had been estimated and was not measured directly. Ice-crystal sizes consisting of less than 1070 of the sample were difficult to obtain due to the program characteristics of the software, which limits the hold increments to O.I°C. Thus, a more complete analysis of the sample (>I 10% ice) was not possible. In order to test for the possible non-specific activity of other hemolymph components, samples of E. solidaginis and of mouse and pigeon plasma were tested for antifreeze activity. Microscopic examination and DSC analysis revealed no activity (Table I; Figs. 2 and 3). The activity of the antifreezes is in agreement with previously reported work which suggested an exponential increase in activity with decreased ice-crystal size (Figs. 4 and 5) [16]. Only the AFP-containing hemolymph revealed a delay in freeze onset. The pigeon plasma did reveal a delay at 15% ice, but not at lower ice-crystal size. All other samples showed a delay of >_0.08 ° C. The DSC results show higher ice fractions (10-3070) than those commonly reported by microscopic observations (>_ 17o) [16] along with high activity. The estimate of ice volume/content in a light microscope is not functionally quantifiable due to (1) resolution limitations and (2) surface area/volume relationships. It is not reasonable to assume an ice-crystal to be a perfect geometrical shape (sphere or square). The DSC, on the other hand, gives an accurate estimate of ice present which is observer-independent. Part of the observed activity difference between the light-microscope and DSC may also be due to the higher cooling rate used by the DSC. Higher cooling rates were observed to result in higher activity in the DSC (data to be presented). The results obtained in agreement with the proposed mechanisms for AFP activity: the smaller the ice-crystal the higher the activity due to a combination of increased AFP available for binding to the crystal as well as decreased ice-surface areas of preferred growth [7,15]. This would also
help explain the difficulty in achieving consistent activity using cold-stage light microscopy, which is dependent on a less precise control of the icecrystal size. Accurate measurements of the ice fraction were possible by using the DSC. Microscopic examination of samples is a rapid test for the presence of antifreeze activity. DSC, however, produces a quantitative analysis that takes into account the amount of ice present in the sample. Acknowledgements This work was supported in part through grants from the Perkin-Elmer Corporation and through NSF (Research G:ant PCM81-10327). References 1 DeVries, A.L. (1971) Science 172, 1152-1155. 2 Duman, J.G. (1979) J. Comp. Physiol. 131, 347-352. 3 Patterson, J.L. and J.G. Duman (1978) J. Exp. Biol. 74, 37-45. 4 Scott, G.K., Fletcher, G.L. and Davies, P.L. (1986) Can. J. Fish. Aquat. Sci. 43, 1028-1034. 5 Kao, M.H., Fletcher, G.L., Wang, N.C. and Hew, C.L. (1986) Can. J. Zool. 64, 578-582. 6 Hew, C.L., Slaughter, D., Joshi, S.B., Fletcher, G.L. and Ananthanarayanan, V.S. (1984) J. Comp. Physiol. B 155, 81-88. 7 DeVries, A.L. (1984) Phil. Trans. R. Soc. Lond. B 304, 575-588. 8 Ramsay, J.A. (1964) Phil. Trans R. Soc. Lond. B 248, 279-314. 9 Grimestone, A.V., Mullinger, A.M. and Ramsay, J.A. (1968) Phil. Trans R. Soc. Lond. B 253, 343-382. 10 Patterson, J.L. and Duman, J.G. (1979) J. Exp. Zooi. 201, 361-367. 11 Tomchany, A.P., Morris, J.P., Kang, S.H. and Duman, J.G. (1982) Biochem. 21(4), 716-721. 12 Schneppenheim, R. and Theede, H. (1980) Comp. Biochem. Physiol. 67B, 561-568. 13 Husby, J.A. and Zacharriassan, K.E. (1980) Experientia 36, 963-964. 14 Ramsay, J.A. and Brown, R.H. (1955) J. Sci. Instrum. 32, 372-375. 15 Hirsh, A.G., Williams, R.J. and Meryman, H.T. (1985) Plant Physiol. 79, 41-56. 16 Zachariassen, K.E. and Husby, J.A. (1982) Nature 298 (5877), 865-867.
217
BBA33250
Differential scanning calorimetric analys|s of antifreeze protein activity in the common mealworm, Tenebrio molitor T h o m a s N. H a n s e n and John G. Boast Center of Cryobiological Research, State University of New York, Binghamton, N Y (U.S.A.)
(Received 20 January 1988) (Revised manuscript received 16 May 1988)
Key words: Antifreeze protein: Protein activity: DSC, differential scanning calorimetry; (7". mofitor)
Antifreeze proteins (AFP) are able to inhibit the growth of ice-crystals at temperatures below the equilibrium freezing poin,~ (Tf) of hemolymph. The analysis of AFP activity has commonly involved the use of direct microscopic observation of a sample following inoculation with ice. The resulting activity, defined as the amount of thermal hysteresis observed between Tf and the subsequent rapid growth of ice, has been reported to range up to 7 o C. However, most studies report high level of variation, possibly due to ice-crystal size variability and the presence of non-visible ice nuclei. We describe a new method of analysis of AFP activity using differential scanning calorimetry (DSC). DSC analysis reveals much higher activity., up to 10°C, with less variation observed within a sample, and is not subject to the difficulty of accurate assessment of ice-crystal volume.
introduction Antifreeze proteins (AFP) form unique classes of macromolecules able to inhibit the growth of ice crystals. These proteins have beea isolated in a variety of species, including fish [1], spiders [2] and insects [3]. The fish AFP have been most extensively studied, and at least four different classes of the protein have been reported based on the presence or absence of carbohydrate sidegroups and on amino acid content [46]. All of the antifreeze molecules lower the hysteretic freezing point (Thy), a secondary onset of crystalliza-
Abbreviations: AFP, antifreeze proteias; Tf, equilibrium freezing point; Thy, hysteretic freezing point; Tm, equilibrium melting point; DSC, differential scanning calorimetry Correspondence: J.G. Baust, Center for Cryobiological Research; State University of New York, Binghamton, NY 13901, U.S.A.
tion in the presence of an ice nucleus. The Thy is lowered in a non-colligative manner: the melting point of the solution is changed in a predicable colligative manner while the T,y is lowered below the equilibrium melting point [7]. in a solution devoid of AFP, the equilibrium melting point (Tm ) and equilibrium freezing point (Tf) are equivalent; however, in an AFP-containing solution the Tf cannot be readily defined due to the action of the proteins. The proposed mechanism for AFP activity is based on the glycoprotein in AFP model. These proteins contain may repeating subunits of Ala-Ala-Thr with a disaccharide linked to the threnonmc. The overall structure forms an a-hello cal rod with a polar side-chain cvc~y 4~ A~ a distance which also separates the oxygen atoms in ice. The polar groups are thought to hydrogen bond with the oxygen atoms in the nascent icecrystal, thereby 'poisoning' the lattice [7]. The protein-crystal interaction would increase the ice front's angle of curvature and disrupt the free-en-
0167-4838/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
218
ergy relationships between bulk water and ice. Thus, a lower temperature for the continued growth of the ice front would be required. The first antifreeze activity was observed in the common mealworm, Tenebrio molitor [8]. Studies revealed the activity to be caused by a protein of molecular weight between 10000-12000 and to be inactivated by trypsin [9]. Further work revealed the possibility of multiple peptides and purified forms were found to not contain carbohydrate side-groups [10-12]. The most common methods of studying the AFP require direct microscopic observation of the sample as it is cooled either on a cold stage [13] or in a capillary immersed in an alcohol bath [14]. The fluid sample is inoculated with an ice crystal (either directly or by whole-sample freezing followed by a partial melt) and then slowly cooled until crystal growth is observed. In the absence of AFP, growth is observed almost immediately (~< 0.02* C) upon cooling below Tm. In the presence of the antifreeze, the growth is delayed as the temperature drops until the AFP activity is overcome and an almost explosive growth of long ice spicules is observed. Different scanning calorimetry (DSC) has been extensively used to study the effect of freezing and thawing on biological samples [15]. By utilizing power-compensated DSC which compares an inert reference to the sample, the enthalpic changes in the sample over temperature a n d / o r time can be monitored. Such changes include exothermic and endothermic events as well as glass transitions either during chemical reactions or during state changes. We propose a new method of analysis of AFP activity using DSC. We suggest that this mode of analysis lends a greater quantitative measure by allowing the ice-crystal volume to be directly measured and controlled. Materials and Methods Samples from four animal species were collected and analyzed for antifreeze activity. Hemolymph was collected from common mealworm larvae ( T. molitor)~ which were maintained at 25 ° C under a 16:8 light/dark photoperiod. Hernolymph from the gall fly third instar larvae, Eurosta solidaginis (October field collections from native
populations) was obtained by cuticular puncture. Plasma from adult female mice (C3H/MeMs x 129/J)F1 C31 (23°C, 14:10 L / D ) and from pigeons, Columba lioia (25°C, 16:8 L / D ) was also collected. Solutions of 0.9% NaCI and 20 mM Tris buffer (pH 7.0) containing 0.5 M NaCI were also analyzed. Samples of 0 . 2 / d or less were suspended in oil (Cargille Type B) and examined under a light microscope connected to a cold stage (Clifton Nanoliter Milliosometer, 100 × magnification). The samples were cooled to - 4 0 ° C, and slowly warmed (0.10 C/min). The temperature at which the last ice-crystal disappeared was noted. The samples were then recooled to - 4 0 0 C ano warmed to a temperature that maintained the smallest visible ice crystal (area = 25 #mZ). Samples were then slowly cooled (0.1 o C/min) and the temperature at which ice-crystal growth resumed was noted. A difference between the hold temperature and that of ice-growth (Thy; 'onset' temperature) was taken as the functional definition of antifreeze activity (hold-onset - AFP activity). Additional samples between 1 and 5 #1 were placed in oil-filled aluminium pans for analysis by DSC. The pans were placed in a Perkin-Elmer DSC-7 equipped with an lntracooler II. An empty aluminum pan was placed in the reference cell. The samples were cooled and warmed at 1 ° C/min and the crystallization and melting points noted. Samples were then cooled to - 4 0 " C (10" C/min), held for 5 min, warmed to various partial melt temperatures between - 1 0 and O°C in 0.1*C increments, and held for 5 min at each increment to allow for ice-protein interaction and system stabiliz~tion. The samples were then slowly cooled to - 1 5 o C (1 ° C/min), and the onset and area of the freeze exotherm calculated. The onset was observed as the start of the cwstalhzation exotherm. The area was calculated from the onset point to the point where the heat content returned to pre-crystallization levels. The ice-inoculated freeze exotherm was compared to the melt endothermic area to calculate the percent ice present in the sample. Oil baselines were run before and after sample analysis. A Perkin-Elmer 7700 computer was used to calculate the areas under the curves and the onset temperature of the exotherms. Thermograms were normalized |'or sam-
219
pie weight. Values are presented as mean _+ standard deviation. H
Results
/:6
v
A summary of microscopic observations of the antifreeze activity are presented in Table I. Individual samples were tested multiple times for activity. A wide range of activity was noted in the AFP-containing samples but not in the samples devoid of AFP, even though careful attention to uniform ice-crystal size was attempted. The T. molitor samples were observed to behave in a manner similar to other AFP, showing an explosive growth of long spicules at Thy. In DSC experiments samples were layered between oil for antifreeze activity measurements. Thermograms of a typical 7'. molitor hemolymph sample are presented in Fig. 1. A delay in the onset of the freeze exotherm was observed to increase with the higher hold temperatures. The area under the curves also increased with the higher temperatures, indicating that less ice was present. In order to test for possible effects by other components of the hemolymph, analysis of AFP-free gall fly hemolymph and of plasma samples was undertaken. Thermograms obtained from E. solidaginis hemolymph are presented in Fig. 2. There were no delay in the onset of the freeze exotherm. Mouse plasma, 0.9% NaCI and Tris buffer showed shailar results to that of the gall fly, with no delay in :he freeze exotherm when ice was
/: _9.o
"-4 a
-1~4
I
-12
I
-10
-8
.~
I
-4
.
2
0
Ternperoture (C) Fig. 1. DSC thermograms of T. mofitor hemolymph illustrating AFP activity. The sample was frozr~,n and allowed to undergo a partial melt before refreezing. The three temperatures presented are - 2.3, - 2.2 and - 2.1 ° C. The - 2.10 C hold (point H) and exotherm onset (point O) are marked. The delay in freeze onset increased as the te~nperature increased. The sampie's crystallization temperatu~'e was - 23.3 5:1.5 ° C (n = 3).
present in the sample. Thermograms obtained from pigeon plasma are presented in Fig. 3. The thgrmograms were similar to the other samples devoid of antifreeze. The oil baselines showed no thermal activity. When a single T. molitor sample was run multiple times to the same hold temperature, some variation in the delay of the exotherm was observed ( - 1 . 2 1 _+ 0.17°C; range - 1 . 0 3 to - 1.50°C; n =
2,0
.?
~ 1.0
TABLE I MICROSCOPIC EXAMINATION OF SAMPLES F O R ANT I F R E E Z E ACTIVITY All samples were analyzed using a cold stage connected to a light microscope and were run multiple times upon inoculation with the smallest ice-crystal observable at 100x(25 #me).
"r
0.0
, -14
, -12
-b
.
~
Temperature
Sample ( . ) a
Hold-freeze_+ S.D. ( ° C )
Range ( * C )
T. molitor (8) T. mofitor (5)
0.70 + 0.21 0.98 + 0.29
0.40-1.01 0.29-1.27
Pigeon Plasma (3) Mouse Plasma (3) Tris Buffer (3)
0.02 + 0.01 0.02 + 0.01 0.01 + 0.01
0.01-0.02 0.01-0.02 091-0.02
" Number of times same sample was tested for antifreeze activity.
.~,
.~
-~
-
o
(C)
Fig. 2, DSC thermograms of E. solidaginis hemolymph illustrating the behavior of a solution devoid of AFP. The sample was frozen and partially melted ( - 2 . 3 , - 2 . 2 and - 2 . 1 ° C ) before being refrozen. The - 2 . 1 ° C hold and exotherm onset are marked by an arrow. The melt exothermic area increased as the temperature increased, indicating less ice was present. The freeze onset was immediate upon cooling in the presence of ice. The sample crystallization temperature was - 17.8+0.9 (n = 3).
220 3,0
12u 10
~8
.-. 2.0
f,
' 6
& O & a
•
O
o ,1
_~1.0
OC
.... ~
,,•
"~2
I
"10
I
I
"8 "6 Temperoture (C)
I
I
"4
"2
,,,
0
Fig. 3, DSC thermograms of pigeon plasma. The sample was frozen and partially melted ( - 3.0, - 2 . 0 and - 1 . 4 ° C) before being refrozen, The hold and onset exotherm for the - 1 . 4 ° C point are marked by an arrow (point S: shoulder on the exotherm). The sample crystallization temperature was - 1 5 + 0.33 = C (n -- 3).
6). The variability was however only a fraction of that obtained by microscopic analysis. The DSC data for T. molitor hemolymph are presented in Fig. 4. The amount of AFP activity and the amount of ice melt are plotted against the hold temperature. There was an increase in the amount of sample melted with increased temperature. There was a parallel increase in AFP activity also with increased temperature. Two hold temperatures ( - 0 . 9 and - 0 . 7 ° C) showed increased melt over adjacent points. The reason in unexplained, but the AFP activity followed the points. "U i
- 8o
10-
60
i
40 E #
20
o <
0
O_ -3
-2
-I
0
HOld (*C)
Fig. 4. T. molitor. Antifreeze activity and per cent sample melt versus actual hold temperature. The amount of sample that melted increased with higher temperature. AFP activity (calculated by subtracting the hold/annealing temperature from the onset of the freeze exotherm) also increased with higher temperature. Two points ( - 0 . 9 and - 0 . 7 ° C ) had higher melts than adjacent points, but also revealed higher activity.
% Ice
Fig. 5. Antifreeze activity versus per cent ice in the ~ample. Samples not containing any antifreeze activity (Tris buffer (~), 0.9% NaCI (<>), and Pigeon Plasma (n) had a difference between the hold and onset of no more than 0.1 ° C. Samples containing antifreeze activity (7". molitor: 0 , &) had a high difference between the hold temperature and the onset of the exotherm. The pigeon plasma difference increased at 15% ice due to a shoulder on the exotherm curve.
All the DSC data are summarized in Fig. 5. The calculated onset was subtracted from the hold temperature and the difference plotted against the percent ice present in the samples (as indicated by the area under the curves). None of the samples exhibited any AFP activity except for the 7". molitor hemolymph. The pigeon plasma showed a delay in freeze onset at 15% ice but not show any delay at lower ice content. An inspection of the thermograms revealed an inflection pnint midway through the freeze, wlfich the computer interpreted as the freeze onset (Fig. 3). Discussion
The microscopic examination for AFP activity revealed a large range of values for activity of repeated individual samples and between samples containing the protein, even when the ice-crystals were carefully monitored for 'constant size' (Table I). A range of activity values has been reported by others [1]. Samples devoid of AFP were, however, found to be very constant in ice-crystal growth characteristics. Analysis of AFP activity by DSC was conducted in a manner similar to the microscopic examinations. Samples containing AFP showed a delay in the freeze onset in the presence of ice (Fig. 1). The delay was observed to increase in an almost exponential manner. Similar results of
22|
increased activity with decreased crystal size have been reported [16], although ice-crystal fraction had been estimated and was not measured directly. Ice-crystal sizes consisting of less than 1070 of the sample were difficult to obtain due to the program characteristics of the software, which limits the hold increments to O.I°C. Thus, a more complete analysis of the sample (>I 10% ice) was not possible. In order to test for the possible non-specific activity of other hemolymph components, samples of E. solidaginis and of mouse and pigeon plasma were tested for antifreeze activity. Microscopic examination and DSC analysis revealed no activity (Table I; Figs. 2 and 3). The activity of the antifreezes is in agreement with previously reported work which suggested an exponential increase in activity with decreased ice-crystal size (Figs. 4 and 5) [16]. Only the AFP-containing hemolymph revealed a delay in freeze onset. The pigeon plasma did reveal a delay at 15% ice, but not at lower ice-crystal size. All other samples showed a delay of >_0.08 ° C. The DSC results show higher ice fractions (10-3070) than those commonly reported by microscopic observations (>_ 17o) [16] along with high activity. The estimate of ice volume/content in a light microscope is not functionally quantifiable due to (1) resolution limitations and (2) surface area/volume relationships. It is not reasonable to assume an ice-crystal to be a perfect geometrical shape (sphere or square). The DSC, on the other hand, gives an accurate estimate of ice present which is observer-independent. Part of the observed activity difference between the light-microscope and DSC may also be due to the higher cooling rate used by the DSC. Higher cooling rates were observed to result in higher activity in the DSC (data to be presented). The results obtained in agreement with the proposed mechanisms for AFP activity: the smaller the ice-crystal the higher the activity due to a combination of increased AFP available for binding to the crystal as well as decreased ice-surface areas of preferred growth [7,15]. This would also
help explain the difficulty in achieving consistent activity using cold-stage light microscopy, which is dependent on a less precise control of the icecrystal size. Accurate measurements of the ice fraction were possible by using the DSC. Microscopic examination of samples is a rapid test for the presence of antifreeze activity. DSC, however, produces a quantitative analysis that takes into account the amount of ice present in the sample. Acknowledgements This work was supported in part through grants from the Perkin-Elmer Corporation and through NSF (Research G:ant PCM81-10327). References 1 DeVries, A.L. (1971) Science 172, 1152-1155. 2 Duman, J.G. (1979) J. Comp. Physiol. 131, 347-352. 3 Patterson, J.L. and J.G. Duman (1978) J. Exp. Biol. 74, 37-45. 4 Scott, G.K., Fletcher, G.L. and Davies, P.L. (1986) Can. J. Fish. Aquat. Sci. 43, 1028-1034. 5 Kao, M.H., Fletcher, G.L., Wang, N.C. and Hew, C.L. (1986) Can. J. Zool. 64, 578-582. 6 Hew, C.L., Slaughter, D., Joshi, S.B., Fletcher, G.L. and Ananthanarayanan, V.S. (1984) J. Comp. Physiol. B 155, 81-88. 7 DeVries, A.L. (1984) Phil. Trans. R. Soc. Lond. B 304, 575-588. 8 Ramsay, J.A. (1964) Phil. Trans R. Soc. Lond. B 248, 279-314. 9 Grimestone, A.V., Mullinger, A.M. and Ramsay, J.A. (1968) Phil. Trans R. Soc. Lond. B 253, 343-382. 10 Patterson, J.L. and Duman, J.G. (1979) J. Exp. Zooi. 201, 361-367. 11 Tomchany, A.P., Morris, J.P., Kang, S.H. and Duman, J.G. (1982) Biochem. 21(4), 716-721. 12 Schneppenheim, R. and Theede, H. (1980) Comp. Biochem. Physiol. 67B, 561-568. 13 Husby, J.A. and Zacharriassan, K.E. (1980) Experientia 36, 963-964. 14 Ramsay, J.A. and Brown, R.H. (1955) J. Sci. Instrum. 32, 372-375. 15 Hirsh, A.G., Williams, R.J. and Meryman, H.T. (1985) Plant Physiol. 79, 41-56. 16 Zachariassen, K.E. and Husby, J.A. (1982) Nature 298 (5877), 865-867.