26, 383-388 (1989)
Scanning Calorimetric Analysis of Tenebrio Antifreeze Protein Activity THOMAS N. HANSEN
JOHN G. BAUST of New York Binghamton,
New York 13901
Recently a new method for analysis of antifreeze proteins by differential scanning calorimetry has been developed (T. N. Hansen and J. G. Baust, Biochim. Biophys. Acta 957, 217-221). However, the parameters used were not examined for possible maximal activity. To test the parameters, pooled hemolymph samples of the common mealworm larva, Tenebrio mofitor (25°C 16 hr:8 hr L:D), were collected and analyzed for activity. The samples were held at -40°C and at various annealing temperatures for different lengths of time (0 to 360 min). No difference in activity was observed in the freeze intervals, while significant differences were observed in annealing times of less than 3 min. Hemolymph samples were also tested for antifreeze activity at various scan rates (0.1 to 10°C . mm’). Significant differences in activity were observed for each rate. Both the short annealing times and the cooling rates were found to be due to methodology and not sample. The best parameters were found to consist of a 5-n& freeze at -WC, a 5-min annealing interval, and a 1°C . min- ’ cooling rate. To test the optimized parameters, pooled samples of T. molitor hemolymph were monitored for changes in activity over time (up to 60 days) at various storage temperatures (- 17, - 80, - 196°C). No changes in activity were observed. These results suggest that care must be given to the reporting of the specific conditions used in the analysis of antifreeze activity. o 1989 Academic press, 1nc.
Antifreeze proteins (AFP) have been discovered and studied in a wide variety of species from antarctic fish (4) to insects (6). These proteins are unique biomolecules that lower the freezing points of water in a noncolligative manner, while the melting point is lowered in a colligative manner (12). The fish antifreeze glycoproteins (AFGP) have been extensively studied. Proposed mechanisms for their activity suggest that the protein interacts directly with the crystal lattice, preventing ice crystal growth in the preferred axis (2, 12). The AFGP consist of eight molecular sizes based on their gel electrophoretic mobilities (3). The higher molecular weight AFGP 1-5 have more activity than the lower molecular weight AFGP 6-8 (8). It has been suggested that the low molecular weight proteins cannot interact with the ice crystal as strongly as the higher molecular weight proteins (12). Received July 14, 1988; accepted January 16, 1989.
Most analyses of antifreeze protein activity are accomplished by microscopic examination of comparatively large crystals (225 p,rn’; 5, 1). Whether the observations are made on a cryostage or in a capillary tube, the investigator controls the temperature of the sample and visually inspects for thermal hysteresis. Cell-free blood samples are inoculated with an ice crystal and then slowly cooled until ice crystal growth is observed. In the presence of AFP, ice growth occurs after a delay of a few degrees (a hysteresis), while in the absence of AFP growth is observed within 0.02”C. Reports on hydrated systems indicated that a “time/temperature dependency” factor may be involved in the freezing of a system (12). The studies suggested that a sample may contain different spheres of water (bulk, perturbed, and unfreezable) and that freezing the sample for various amounts of time or to different temperatures may result in more or less of the perturbed water entering the crystalline state. The studies suggested that investigators
383 OOII-2240/89 $3.00 Copyaht 8 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.
should determine whether a time/temperature phenomenon is present in their system and as such report the freezing conditions used. Also, AFP-containing systems have not been tested for possible activity alterations after multiple freeze-thaw cycles. Recently, a new method for the analysis of AFP by differential scanning calorimetry has been detailed (5). We now report on the further development of this method to involve parameters designed to test maximal activity and sensitivity of AFP analysis. Pooled hemolymph samples of the common mealworm, Tenebrio molitor, were tested for changes in activity with various freeze and annealing times as well as cooling rates. To test the optimized parameters, hemolymph samples were monitored for changes in activity over time at various storage temperatures. MATERIALS
Larvae of the common mealworm, Tenebrio molitor, were maintained at 25°C (16 hr:8 hr 1ight:dark photoperiod). Hemolymph samples were obtained by cuticular puncture, pooled in capillaries between layers of oil, and sealed. The samples were analyzed by differential scanning calorimetry (DSC) as previously reported (5). Briefly, 5 ~1 of freshly pooled hemolymph was put in an oil-lined 20-p,l aluminum pan, weighed, and loaded into a Perkin-Elmer DSC-7. The sample was cooled to -40°C warmed to various annealing temperatures (between -4 and 0°C in O.l”C increments), and retooled to - 40°C. The onset and area of the exotherm were noted for each temperature. The sample was also scanned at 1°C . min-’ between + 10 and -40°C to record the crystallization temperature, melting point, and area. Antifreeze activity was defined as the difference between the annealing temperature and the onset of the freeze exotherm: annealing temperature - onset temperature = AFP activity (“C).
When maximal AFP activity is reported for the DSC, the “annealing temperature” is similar to the “melting point” of other analysis techniques. The various parameters of the DSC analysis procedure (tested at the same annealing temperature) included: (1) freeze times (-40°C): 0, 2.5, 5, 15, 30, 60, 180, and 360 min; (2) annealing times: 0, 1, 3, 5, 15, 30, and 60 min; and (3) cooling rates: 0.1, 0.5, 1, 2.5, 5, and 10°C * min-‘. All conditions were replicated three times. To control for temperature calibration changes in the DSC with the various cooling rates, a water sample was prepared as above and run through multiple freeze-thaw cycles (n = 9), and the onsets of the freezes and melts were noted. Values are presented as means 2 SD. Results were analyzed statistically by ANOVA (10) and by student NewmanKeuls (SNK) analysis (10) for a breakdown of any variation within a set of conditions. To test the optimized parameters, pooled hemolymph samples were stored at various temperatures (- 17, -80, - 196°C) and monitored for activity over time (0, 10, 30, 60 days). RESULTS
Analysis of the pooled T. molitor hemolymph revealed typical insect antifreeze activity; an increase in activity was observed with decreased ice content of the sample (Fig. I). The maximal activity observed was a delay in the exotherm onset of 2.02”C, with 10.7% ice present. A best-fit exponential curve for the sample was y = 2.12 x lO-o.o12x, R = 0.85. Once the pooled hemolymph was well characterized, it was used to test the various parameters of the analysis program in an attempt to optimize the system. To determine whether such phenomena as freeze time and temperature could alter the AFP activity of T. molitor hemolymph, a sample was held at -40°C for various amounts of time (Fig. 2). No significant difference in the exotherm onset in the pres-
F .5 03
-z a, POI
Temperature (OC) FIG. 1. Representative thermogram of antifreeze activity. The sample shown consists of pooled T. molitor larvae hemolymph that was stored for 60 days at - 17°C. The three cooling ramps shown (from top to bottom) are for annealing temperatures of - 1.6, - 1.5, and - 1.4”C. The exotherm onset and amount of ice present for each temperature are noted by the individual curves. The sample’s mean crystallization temperature after supercooling was - 25.41”C 2 4.31. Endo, endothermic.
ence of ice was observed (ANOVA, F(6,14) = 0.608; P 3 0.25). This suggested that there is no freeze-time factor involved in AFP activity. However, the analysis used to check for time dependency was an alteration of the methods used for such studies: The sample is usually held at a specific condition and then warmed through the melt,
and the area of the melt is calculated. If a time dependency factor does exist, then the melt endotherm would be different (11). To test for time/temperature dependency according to the original method, the sample was held at various temperatures (-40,
c t r : 0"
. 8 I
I'I,."I 3 5
z -2.25 -2.50
I"I"--f 30 60
Log Time (min)
FIG. 2. T. molitor antifreeze protein activity after various freeze times. No significant difference in the exotherm onset in the presence of ice was observed (ANOVA F(6,14) = 0.608; P 2 0.25).
FIG. 3. T. molitor antifreeze protein activity after various annealing times. ANOVA analysis revealed significant differences in the exotherm onset between conditions (F(6,14) = 231.909; P G O.OOl),which SNK analysis revealed to be for annealing times of less than 3 min (P < 0.001; Table 1).
HANSEN AND BAUST TABLE 1 SNK Analysis of the Various Annealing Times
Annealing time (min.) Mean onset (“C)
60 - 1.84*
3 - 1.84*
0 - 2.39
Note. The mean onset temperature is presented below the annealing time. Values marked with an asterisk were not significantly diierent at the P < 0.001 level. ANOVA F(6,14) = 231.909; P S 0.001.
- 100, - 170°C) for 5 min and warmed through complete melting. No significant difference in the melt area was observed (ANOVA, F(2,6) = 0.305; P > 0.25). The results were not altogether unexpected as the hemolymph is a dilute solution, whereas the systems used in the time dependency studies consist of hydrated solids. A freeze time/temperature factor could alter AFP activity, but the change in signal observed upon melting would be masked by the large bulk-water endotherm of the highly diluted solution. The results also suggest that multiple freeze-thaw cycles do not alter AFP activity, although this was not directly tested. The AFP was next tested for any changes in activity after various annealing times
(Fig. 3). ANOVA analysis revealed a significant difference between the conditions (F(6,14) = 231.909; P s 0.001). SNK analysis showed that annealing times of less than 3 min were significantly different (P < 0.001; Table 1) but not annealing times of 3 min or longer. Inspection of the thermograms revealed that the DSC was not yet stable when cooled before 3-min holding (Fig. 4). The freeze exotherm areas for annealing times above 3 min were not significantly different (ANOVA, F(4,lO) = 1.007; P > 0.25), indicating that the ice crystal structure was stable. Annealing times for the AFP could have affected the onset of the freeze exotherm by allowing the ice crystal to continue melting, resulting in higher activity. However, the structure was
FIG. 4. DSC thermograms for 0- to 3-min annealing times. The thermograms are presented on a time scale, where the end of the annealing times are indicated by arrows. The calorimeter required 3 min to stabilize, resulting in an artifactual increase in the exotherm onset for the lower annealing times. Endo, endothermic.
found to be stable (the percentage ice remained constant) even after 60 min. Others have reported that there is no change in ice crystal size after several days annealing (7). Annealing times of less than 3 min did reveal an artifactual increase in activity due to instrument instability prior to cooling (Fig. 4). The AFP was next tested for changes in the freeze exotherm onset at different cooling rates. The sample was observed to have significantly higher activity with higher cooling rates (Fig. 5; ANOVA, F(5,12) = 258.81; P 5: 0.001). A water sample was prepared to test whether the change in activity was due to the sample or to an instrument factor. The onset of the water crystallization point was calculated for each scan rate. The data were then normalized to 0°C for the 0. 1°C s min- * scan and compared to the hemolymph sample. The hemolymph onsets in the presence of ice were observed to parallel the freeze exotherm onsets of the water sample, indicating that the AFP ac-
I ’ “.I
1 .o Rate
m ICC FIG. 6. T. molitor antifreeze activity after 60 days of various storage conditions. No difference in activity was observed for storage conditions at - 17°C (0), -80°C (A), and - 196°C (m) as compared to a fresh sample (Cl). Similar results were observed after 10 and 30 days of storage.
tivity is independent of cooling rate. These results are in agreement with previous researchers who had observed AFGP 1-5 to be cooling rate independent (8). These results indicate that caution must be taken when analyzing a sample on the DSC since higher cooling rates appear to yield higher antifreeze activity. The results suggested a preferred protocol of a 5-min hold at -40°C and at the annealing temperature to allow for machine stabilization. A 1°C * min-’ cooling rate was also chosen. These parameters were used to test pooled hemolymph after 60 days of storage at - 17, -80, and - 196°C. No difference in activity was observed for the samples as compared to the fresh control (Fig. 6).
FIG. 5. T. molitor antifreeze activity protein for various scan rates. Onset of crystallization points (0) of a water sample, as a function of scan rate, compared to pooled T. molitor hemolymph (m) cooled in the presence of ice. The O.l”C . min-’ onset points have all been shifted to 0°C. The observed decrease of the freeze exotherm onset value with increased cooling rate was noted to be an instrument factor. ANOVA analysis revealed the conditions for the hemolymph sample to be signiticantly different (F(5,12) = 258.81; P < 0.001). The apparent increase in antifreeze activity of the hemolymph parallels the increase in the mean freeze exotherm onset.
1. DeVries, A. L. Glycoproteins as biological antifreeze agents in antarctic fishes. Science 172, 1152-l 155 (1971). 2. DeVries, A. L. Role of glycopeptides and peptides in inhibition of crystallization of water in polar fishes. Philos. Trans. R. Sot. London B 304, 575-588 (1984). 3. DeVries, A. L., Komatsu, S. K., and Feeney, R. E. Chemical and physical properties of freezing point-depressing glycoproteins from antarctic fishes. J. Biol. Chem. 245, 2901-2908 (1970).
4. DeVries, A. L., and Wohlschag, D. E. Freezing resistance in some antarctic fishes. Science 163: 1073-1075(1%9). 5. Hansen, T. N., and Baust, J. G. Differential scanning calorimetric analysis of antifreeze protein activity in the common mealworm, Tenebrio moliror. Biochim. Biophys. Acta 957, 217-221 (1988). 6. Ramsay, J. S. The rectal complex of the mealworm Tenebrio molitor, L. (Coleoptera, Tenebrionidae). Philos. Trans. R. Sot. London E 248, 279-314 (1964). 7. Raymond, J. A., and DeVries, A. L. Adsorption inhibition as a mechanism of freezing resistance in polar fishes. Proc. Natl. Acad. Sci. USA 74, 2589-2593 ( 1977). 8. Schrag, J. D., and DeVries, A. L. The effect of freezing rate on the cooperativity of antifreeze
IO. I I.
glycopeptides. Comp. Biochem. Physiol. A 74: 381-385 (1983). Schrag, J. D.. O’Grady, S. M., and DeVries, A. L. Relationship of amino acid composition and molecular weight of antifreeze glycoproteins to non-colligative freezing point depression. Biochim. Biophys. Acta 717, 322-326 (1982). Sokal, R. R., and Rohlf, F. J. “Biometry: The Principles and Practice of Statistics in Biological Research. Freeman, San Francisco, 1969. Wolanczyk. J. P., and Baust, J. G. The influence of hold time on the phenomenon of ‘time dependency’ in hydrated lysozyme glasses. CryoLert. (in press) (1989). Yeh, Y., and Feeney, R. E. Anomalous depression of the freezing temperature in a biological system. Act. Chem. Res. 11, 129-135 (1978).