SOME PROPERTIES OF THE GLUCOSE DEHYDROGENAsE FROM SPORES OF Thesis fot the Dogma of Ph. D« MICHIGAN STATE UNIVERSITY John Alfred Bach 7 ~ ; ‘ 1953- . THESIS This is to certify that the thesis entitled Some Properties of the Glucose Dehydrogenase from Spores of Bacillus cereus presented by John A. Bach has been accepted towards fulfillment of the requirements for wdegl‘ee inflCLQflQlOQy and Public Health @414,“ 4,7] LW Major professor h/ {‘71) Date March 29, l963 O~169 LIBRAR y Michigan State niversity ABSTRACT SOME PROPERTIES OF THE GLUCOSE DEHYDROGENASE FROM SPORES OF BACILLUS CEREUS by John A. Bach This research was undertaken in an effort to under— stand the mechanism of heat resistance in bacterial spores by examining the properties of heat resistant spore proteins. Specifically, this approach involved the purification and characterization of glucose dehydrogenases from sporulating cells and germinated spores of Bacillus cereus. In previous studies these enzymes were identical with respect to their serological and substrate specificities but differed in their levels of thermal stability. However, these differences could no longer be demonstrated in the present investigation. In attempting to account for these discrepancies, the heat resistance of glucose dehydrogenase was examined in both intact and ruptured sporulating cells. dormant spores, and germinated spores. The half—life of the enzyme in each preparation (with the exception of intact dormant spores) was about 5 min at 65 C. The enzyme in dormant spores had a half—life of at least 50 min at 90 C. The discovery was John A. Bach made that slight changes in solvent properties such as pH and ionic strength profoundly affected the stability of this enzyme. The half-life at 65 C of the glucose dehydrogenase in a spore extract was increased to about 7 hr by concentrating the extract. Dialysis of this concentrated extract at constant volume against distilled water resulted in a 125 fold decrease in heat stability but all of the original stability was recovered by adding back the lyophilized dialysable material. A glucose—6—phosphate dehydrogenase in these extracts was also stabilized by the dialysable material but not to the same degree. Glucose dehydrogenase from germinated spores was purified 3.400 fold by a sequence of mechanical extraction, protamine treatment, ammonium sulphate fractionation, and ion exchange chromatography with DEAE cellulose. Some physical—chemical properties of this purified enzyme which might be related to heat resistance were studied in relation to the solvent composition. Heat resistance was found to be strongly dependent on the hydrogen ion concentration, exhibiting a sharp peak at pH 6.5 in 0.05 molar imidazole buffer. The Arrhenius activation constant in this buffer —1 was 17,000 K . p. l . J. . m ..|. .i. . . \. I . . t. I a r. . \ a. . a u .. .. .. r . . . .J . ‘v i r. _. . t f John A. Bach High concentrations of certain ionized compounds including NaCl, KCl, NaZSO4, (NH4)ZSO4, and calcium-DPA chelate also increase the spore enzyme heat stability. The increase in stability in NaCl solution is approximately 2nd order with respect to the concentration of NaCl. The effect of CaCl2 was found to be more complex than the NaCl effect. At concentrations up to 0.8 molar. CaCl2 protects the enzyme from heat inactivation while at concentrations above 0.8 molar CaCl2 this effect is reversed. Purified glucose dehydrogenase is very resistant to guanidine inactivationat O C and the presence of 2.5 molar NaCl increases this resistance. The inactivation reaction is approximately 8th order with respect to the guanidine concentration in both low and high ionic strength solvents. The molecular weight of purified glucose dehydrogenase in a low and high ionic strength solvent was determined by a combination of diffusion and sedimentation measurements. In 0.05 molar, pH 6.5 imidazole buffer. with and without 5 molar Nacl, the molecular weight was approximately 13,000 and 32,000 respectively. Although it is entirely possible that a highly ionic environment is responsible for the heat resistance of this enzyme in intact spores, the evidence obtained to date is =.:ni. E: Z P- L) < u: 2 >. DJ 2 h] Figure l. _i 2 cc m a. «n l: 2: :3 Figure 2. l200 - 800- 400} l l 0 ML. OF ENZYME SOLUTION 0.] 0.2 The dependence of glucose dehydrogenase activity on enzyme concentration. 540 O 530'- ——————“‘——--————i G O 520 - 9 l I l l 0 5 IO l5 TIME IN MINUTES The effect of temperature equilibration time on the activity of equal samples of glucose dehydro— genase. 15 F. Glucose—6-phosphate Dehydrogenase Assay This assay was performed in the same manner as the glucose dehydrogenase assay but 0.5 micromoles of TPN and 2 micromoles of glucose-6-phosphate Were used in place of DPN and glucose. G. Protein Determinations Protein concentrations were determined by the spectro— photometric method of Warburg and Christian (1942). This method is based on the relative absorption at 280 and 260 millimicrons of purified yeast enolase and yeast nucleic acid. In using this method, it is assumed that the proteins being measured have absorption spectra similar to yeast enolase. H. Heat Inactivation Measurements The termal inactivation of proteins is a first order reaction. The rate of inactivation is given by the following expression: 429—) = k(E) (1) where (E) is the concentration of active enzyme, t is time, and k is the inactivation rate constant. The following integrated equation relates the concentration of active l6 enzyme at time t to the initial concentration (E0): (E) 2.303 log ——I%T— = kt (2) The actual concentration of enzyme in terms of protein need not be known since the change in enzyme activity with time also gives a measure of k as follows: A log A0 = kt or (3) log A = —kt + log A.O (4) Equation (4) is the familiar form of a linear equation with slope -k and intercept log A0. It is characteristic of a first order reaction that a constant fraction of the material present at any moment reacts in a given time interval. For this reason the thermal inactivation rate of an enzyme can be expressed in terms of the time required for 50% inactivation or the half—life. The relationship of half—life or t to k is obtained from 1/2 equation (4) as follows: A 0 log Ao/2 or (5) t = log 2 _ .302 (6) 1/2 k k The experimental procedure for obtaining inactivation rates was as follows: small samples of the enzyme solution 17 (0.05 ml — 0.5 ml) were heated in rubber stoppered 13 x 100 mm pyrex test tubes for various times in a water bath. The temperature of the bath could be controlled to 0.01 C. The tubes were submerged almost to the stopper to prevent concentration changes due to refluxing. After heating, the tubes were cooled rapidly in cold tap water and the enzyme activity was assayed. Inactivation constants were obtained from the slope of a plot of log activity versus time according to equation (4). The inactivation rates of enzymes in intact spores or cells were obtained in the same manner except that the heated cell suspensions were extracted prior to assay to release the enzymes. A VirTis mixer (The VirTis Co., Inc.) and number 110 glass beads were used for the extractions. Reproducible results could be obtained by this procedure. L'JJUJ .‘Tn' 5.! .' 7T ' ',"fi'1 RESULTS Glucose dehydrogenase from 73 g (dry weight) of germinated spores was purified in the manner described by Bach and Sadoff (1962). This modification of the procedure of Doi, Halvorson and Church (1959), consisted of mechanical disintegration of spores, high speed centrifugation (32,000 G) of the extract, precipitation of nucleic acids with protamine sulfate, ammonium sulfate fractionation and DEAE cellulose column chromatography. The enzyme obtained repre— sented a 2,160 fold purification with respect to the dry weight of germinated spores. Attempts to crystallize the enzyme by gradual addition of saturated ammonium sulfate (pH 6.0) were not successful. Instead, an amorphous material of 3,400 fold purity was precipitated. Cellulose acetate electrophoresis of this material in the discontinuous buffer system of Goldberg (1959) yielded one broad dark band and one very thin light band when stained with nigrosin. The final yield of purified enzyme was 2.5% of the initial activity. The specific activity of the glucose dehydrogenase is 86 micromoles of DPN oxidized per mg of protein per min and its concentration in germinated spores is 0.03% of the dry weight. 18 19 As stated earlier, the experimental approach to the present study was to compare the properties of a heat stable glucose dehydrogenase from sporulating cells with those of an homologous heat labile glucose dehydrogenase from germinated spores. The heat inactivation half—lives reported for these enzymes (Bach and Sadoff 1962) were 15 min at 70 C and 1 min at 50 C respectively. However, contrary to expectations, the purified germinated spore enzyme had a half—life of 25 min at 60 C. Approximately the same level of stability was found in samples from various steps throughout purification. Attempts to alter the heat resistance of the enzyme in extracts of germinated spores by varying conditions of growth, germination and extraction were not successful. Finally, the in vivo and ig_yit£g heat stability was measured in sporulating cells, dormant spores and germinated spores. With the exception of intact dormant spores, the enzyme in each case had a half—life between 1 and 10 min at 65 C. HoWever, in intact dormant spores, it had a half—life of at least 60 min at 90 C. The glucose dehydrogenase does not appear to be intrinsically stable at any stage of sporu— lation but rather appears to be stabilized by conditions in intact dormant spores. 20 Because of this discovery, the approach to the problem of spore heat resistance was shifted to an analysis of the heat resistance of glucose dehydrogenase in various solvents. An effort was made to reproduce the aqueous environment which must exist in intact dormant spores. Since spores contain a high concentration of calcium, the effect of calcium chloride on the heat stability of glucose dehydrogenase in a spore extract was tested. Solutions of sodium chloride were used as a control of the effect of ionic strength. Both salts increased the enzyme stability at a molar ionic strength of 0.15 but, when the calcium chloride concentration was increased to 1.5 molar, rapid inactivation of the enzyme at room temperature ensued. On the other hand, increasing the NaCl concentration greatly enhanced the heat stability of the enzyme. At a NaCl concentration of 3 molar, the spore enzyme had a half—life of about 8 min at 95 C. The same increase in thermal stability was noted for the glucose dehydrogenase from germinated spores. Other spore enzymes were tested for the effect of NaCl on their heat stability. Although some enhancement of stability was noted, it was not equivalent to that found for glucose dehydrogenase. For example, the heat resistance of catalase and g1ucose—6—phosphate dehydrogenase in spore ml: i._ .- .20 slam a.w flxulie n8 .1. :.::'| jam“ 21 extracts increased only about 4 fold when 3 molar NaCl was added. A variety of compounds were examined to determine whether the increase in heat resistance when NaCl was added was due to a specific or non—specific effect. Salts with monovalent cations and mono— or di—valent anions such as NaCl, KCl, Na2804, and (NH4)ZSO4 at 2 molar concentration and pH 6 increased heat stability as much as 700 fold while salts with divalent cations such as 2 molar CaCl and MgCl 2 2 inactivated the enzyme at room temperature. Sucrose, at a concentration of 1.4 molar and pH 6.0, only doubled the heat stability. Because of the relationship of DPA and calcium—DPA chelate to spore heat resistance, the effect of these compounds on the heat stability of purified glucose dehydrogenase was also determined. Potassium dipicolinate, at a concentration of 0.8 molar and pH of 6.4, increased the heat stability of the enzyme considerably. This effect was enhanced on adding an equal amount of CaCl to the system. Presumably a 2 calcium—dipicolonate chelate formed under these conditions. However, when compared under conditions of equal ionic strength and pH, the chelates were no more effective than DPA alone, although chelation did prevent the enzyme inacti— vation which had been observed with high concentrations of CaClz. A; .'....' Lifflfi ': 3 in a .I‘; hr: 1 Dbl j —l:- l...“- .'—':'::.' .. '. a .2 ' 1.5 ' E _ ’- ‘ I — \ .. \ ' nsflw annujaiauu I . Ll' EILHVC -...‘-.. . Jssd n1 canozoni 5d: TLLI- If . 3A,- L.-J - *.- ~39 niilooqe s o: i_ uni-J; ‘3. sub ' dueisvonon “T .losu -..l. L. 11!; 22 The stability of glucose dehydrogenase in the presence of DPA was particularly meaningful because of the known high DPA content of spores and the extremely stable character of this enzyme in intact spores. It seemed possible that a highly ionic environment was responsible for the heat resistance of this enzyme in intact spores and that the extraction procedure diluted this environment. To test this idea, an extract prepared from 15 g (wet weight) of spores was lyophilized and then resuspended in a small amount of distilled water. After centrifugation to remove insoluble material, about 3 m1 of the extract was recovered. The stability of two enzymes in this extract was determined, the glucose dehydrogenase at 65 C and the glucose—6—phosphate dehydrogenase at 45 C. The extract then was dialysed repeatedly against distilled water and the same two enzymes were examined for heat resistance. Finally, the low molecular weight materials which diffused into the distilled water were recovered by evaporation and added back to the dialysed extracts. The heat stability of the enzymes in this Freconstituted? extract was measured. The reconstituted extract contained double the original concentration of dialysable material because about half of the dialysed extract was used for inactivation measurements prior to Freconstitution.9 23 Figs. 3 and 4 clearly show the stabilizing effect of this dialysable material on the two enzymes. The decrease in heat resistance of the glucose dehydrogenase due to dialysis was 125 fold and all of the original resistance was recovered by adding back the dialysable material. The stability of the glucose—6-phosphate dehydrogenase was only decreased 14 fold and the addition of the dialysable material did not restore all of the original heat resistance. When samples of dialysed and undialysed extract were completely dried before heating, the glucose dehydrogenase in both extracts had a half-life of about 50 min at 90 C. The results described in the preceding paragraph led to a detailed study of the effect of the solvent on the enzyme. The heat stability of the purified enzyme was measured at 55 C in 0.05 molar imidazole buffer at various pH values as shown in Fig. 5. Imidazole was chosen because it buffers well from pH 5.5 to 7.5. In Fig. 6 a plot of the logarithm of the inactivation constants obtained from Fig. 5 versus pH shows that the enzyme stability is at a maximum (minimum k) at pH 6.5 and decreases rapidly on both sides of this value. A similar dependence on pH was observed by Levy and Benaglia (1950) for the thermal denaturation of ricin, and the kinetics of this phenomenon were discussed by '24 >- 2.0 t . . , UNDIALYSEID 2 ° C ° '— l.8 '— o o 2 LG - RECONSTITUTEI *- I 2: 0 u: 0 L4 . 0‘ DIALYSED 0 m l.2 - o. 8 LO - _i I 1 I00 200 HEATING TIME IN MINUTES Figure 4. The heat stability of glucose-6-phosphate dehydro— genase in undialysed, dialysed, and reconstituted concentrated spore extract at 45 C. > 2.0 ,RECONSTITUTED .q__ —th» I: I ' 44- 2 LB ,_ UNDIALYSED t) ( LB '2' u, 1.4 l) 5 I2 0_ , DIALYSED 01.0 - 3 . 1 0 I00 200 HEATING TIME IN MINUTES Figure 3. The heat stability of glucose dehydrogenase in un— dialysed, dialysed, and reconstituted concentrated spore extract at 65 C. LOG OF PER CENT ACTIVITY Figure 5. 2.0 O @ °~T~TTNEL- pld 5.5 l.2 - pH 7.5 0 IO 20 30 40 HEATING TIME IN MINUTES pH 6.0 pH 5.5 I I l O ICC 200 300 HEATING TIME IN MINUTES The effect of pH on the heat inactivation of purified glucose dehydrogenase at 55 C. -3.0- \SLOPE=2.5 - 4 .0 l l l I 5,0 6.0 7.0 8.0 pId Figure 6. The effect of pH on the heat inactivation of purified glucose dehydrogenase at 55 C. 27 Scheraga (1961). He attributed the changes in stability to Changes in the ionization of groups involved in side—chain hydrogen bonding. Since pH is defined as the negative logarithm of the hydrogen ion concentration (or activity), Fig. 6 is acutally a plot of log k versus log (H+). This situation is described by the following equations: k = a(H+)n or (7) log k = n log (H+) + log a _ (8) In equation (8) the slope n indicates the order of the reaction with respect to hydrogen ion concentration. Assuming that side chain hydrogen bonds are being broken during denaturation, the values of the slopes in Fig. 6 are indicative of the number of ionizations involved in the stability changes above and below pH 6.5. The scatter observed in the heat inactivation data constitutes errors of 20 to 30%; much higher than the usual 1 or 2% sampling errors. These errors suggested that the enzyme was being reversibly denatured and that the effect on the apparent enzyme stability was not being taken into account. To test this idea, samples of enzyme solution at pH 5.5 were heated 15 min at 55 C and assayed immediately and after a 5, 30, and 90 min incubation at 25 C. A 40% increase in enzyme activity occurred during the first 5 min 28 of incubation and this increased to 60% after 30 min, indi— cating a reversal of thermal denaturation. On the other hand, the same type of experiment at pH 6.5 gave no increase after an 83 min incubation at 25 C. Apparently the renatur— ation only occurs at the lower pH values. Because the heat stability of glucose dehydrogenase is so sensitive to changes in solvent composition, the inactivation constants in various solvents may differ greatly at the same temperature. This necessitates that two separate temperatures be employed to obtain measurable activity dif— ferences in a reasonable time interval. Thus, some means of comparing the inactivation constants at different temper— atures is useful. The relationship of temperature to inacti— vation constant was determined empirically by Arrhenius as: k = Ae-a/T or (9) 1n k = -a/T + b (10) The constant a or activation constant is a measure of the temperature dependence of the rate of a reaction and is given by the slope of a plot of log k versus l/T where T is the absolute temperature. Fig. 7 shows such a plot for the inactivation of purified glucose dehydrogenase in imidazole buffer at pH 6.5. The activation constant obtained from the 5M.CH3E::IZCHDO -2LC)- 1 . 1 1 1 280 290 300 BIO VT x I05 Figure 7. An Arrhenius plot for the inactivation of purified glucose dehydrogenase at various temperatures in 0.05 molar, pH 6.5 imidazole chloride buffer. 30 slope of this curve is 17,000 OK- . If the activation constant a and the inactivation constant k at a given temper— ature are known, it follows that the k at any temperature can be calculated from the following equation: 1n k2 - 1n k1 = -a T— — T— (11) Although the Arrhenius equation has been given a fuller interpretation by modern reaction rate theories, the original empirical equation is adequate for the purpose of comparing inactivation constants at different temperatures. The effect of ionic strength on the thermal resistance of glucose dehydrogenase was studied more quantitatively in an effort to elucidate the mechanism of protection. Enzyme inactivations were performed at 85 C in a pH 6.5, 0.05 molar imidazole chloride buffer with various concentrations of NaCl. In Fig. 8, the log of k is plotted against the log of the NaCl concentration in a manner analogous to the pH dependency curve in Fig. 6. The slope of the curve in Fig. 8 is —2.3 suggesting that the protective reaction involves the binding of two Nafions or two CIIions or two of each. It is of interest that the slopes of Fig. 6 and Fig. 8 are so similar. This could simply be a coincidence or it might indicate a similarity in the mechanism of stabilization by LOG OF k Figure 8. 31 SLOPE = -2.3 I 0.0 LOG OF Na CI CONCENTRATION The effect of NaCl concentration on the thermal stability of purified glucose dehydrogenase at 85 C. In addition to the NaCl concentrations indicated, the solvent contained 0.05 molar imidazole chloride buffer at pH 6.5. |.0 32 changes in hydrogen ion and NaCl concentration. The signs of the slopes are opposite, of course, because the increasing pH values in Fig. 6 correspond to decreasing concentrations of hydrogen ions. The same type of experiment was performed to determine the effect of the calcium chloride concentration on the heat stability of glucose dehydrogenase. In this case, a pH 6.0, 0.05 molar imidazole buffer was used because this was thought to be the pH of maximum stability at the time this experiment was performed. As seen in Fig. 9, the enzyme is most stable at a CaCl2 concentration of 0.8 molar and loses stability with increasing concentration. It should be noted that the kinetics of the CaCl2 effect are not described by equation (8) since the curves are not linear. Stewart and Halvorson (1954) and Black and Gerhardt (1962) have suggested that enzymes in spores are stabilized by intermolecular polymerization. To determine if polymeri— zation was responsible for the heat stability of glucose dehydrogenase, the molecular weight of the enzyme was deter— mined in both a low and high ionic strength solvent by a combination of diffusion and sedimentation measurements. A very small amount of glucose dehydrogenase protein was available so the usual methods utilizing refractive index Figure 9. 33 - I.O 0.0 LOG OF Ca Ce;t CONCENTRATION The effect of calcium chloride concentration on the thermal stability of purified glucose dehydrogenase at 75 C. In addition to the CaCl concentrations indicated the solvent contained 0.05 molar imidazole chloride buffer at pH 6.0. 34 measurements could not be used. Instead, methods involving enzymic activity were employed to take advantage of the very sensitive assay for this enzyme. Diffusion measurements were made in the Stokes cell pictured in Fig. 10. In this device, material at a certain concentration in compartment A (lower compartment) is trans— ferred by diffusion through the fritted glass filter into pure solvent in compartment B (upper compartment) under the driving force of a concentration gradient located within the filter. Independent mixing by the magnetic stirring bars maintains a uniform concentration in both compartments but does not disturb the gradient in the fritted glass filter. The entire cell was immersed in a water bath of i 0.02 C accuracy, to prevent disturbances from convection and volume changes. The mass transfer in the cell is described by Ficks first law of diffusion as follows: dc dm — —DA dx dt (12) where m is the mass of material transferred in time t, A is the area of the membrane, dc/dx is the concentration gradient across the membrane, and D is the diffusion coefficient characteristic of the material under study. If diffusion continues for an appreciable time, the concentrations of IN S -MAGNET \ STIRRING BAR RUBBER STOPPER ,TEFLON STOP COCK Figure 10. The Stokes diffusion cell. 36 material in each compartment will change significantly and the integrated form of equation (12) must be used. This is given as: 1 03 CS D = Bt 1n t t (13) Cl C u where CE and C: are the concentrations of material in the lower and upper compartments respectively at time 0, CE and c: are the concentrations at time t, and S is the cell constant which is equal to: A l 1 ‘3‘ 2’ VJ Vu The symbols vu and vz represent the volumes of the upper and lOWer compartments respectively, and if is the effective thickness of the porous plate. The compartment volumes were obtained from the weight of water they contained at 25 C divided by the known density of water at this temperature. The upper compartment held 22.84 ml, the lower compartment 23.79 ml, and the fritted filter 0.40 ml. The ratio A/Q' or the cell constantfi can be determined from equation (13) by using a material with a known diffusion constant. A 0.1 molar solution of KCl in water was used for this purpose. The diffusion constant of this solution at 17.5 C is 1.38 cmZ/day 5 — 2 or 1.60 x 10 cm /sec (Handbook of Chemistry and Physics, E42" ‘3 ,‘J um}: :fis'. "'_ .:'==.7j.f,‘:-r.'.qz:r=,: -:.+r;nm:,".:sr_-mo', aeqqu bur-z Iml . .‘l .':.;r..i.:'.". :ln-n: rtr.-. orb" M35 1 D 1 : L I .. I. . . fir. ‘u f ".- :.' ,'I= e . min-1w S. '. . a . » l \ . \ ~4 ) . I - ~ \ I ... L‘ - - . 37 40th ed.). Since the cell was calibrated at 15 C, the following relationship between diffusion coefficient and absolute temperature was used to obtain D of KCl at 15 C: where’l is the viscosity of water and the subscripts 1 and 2 refer to two different experimental temperatures. The 5 cmZ/sec diffusion coefficient of 0.1 molar KCl is 1.49 x 10— at 15 C. The experimental procedurefor obtaining diffusion rates was as follows: All solutions were first degassed by boiling at atmospheric pressure or under high vacuum. The entire cell was then filled by suction with the solution under study and placed in the water bath for temperature equilibration. After equilibration the top compartment was emptied, rinsed, and refilled with pure solvent. The magnetic stirring device was set at 120 rpm and diffusion was allowed to proceed for an appropriate period of time. At time t, the solution in the upper compartment was removed and replaced with fresh solvent to begin another period of diffusion. This procedure was repeated until constant values 0 o l QL — Cu of the diffusion function —E— 1n “~E*———E— were obtained; Call-CU. indicating that the concentration gradient dc/dx in the "IT-._“" LI: ' 1. £5 sin his” uni-sir" _ _. .~ - aerffi . u I ‘1 ! . t'i‘r-M' E but 1 asqixoaduu ad: has .ejsa 90 fififlfnliv on: at am .133: ::--=:-.-'.'~r.mn:.- -s;-r:";n.' . r 1'" .' r'..'23'3'r.t.- 0w: 0: 1.19! i" 1.‘ '2". if. .’ - .. t...' .133? ' r‘...*.'. ..‘ .' .1." " f-‘fi-"I': fl'.‘ “13116 , . . 'n t L in J-J rt ’ '-' ' b- 38 membrane had reached a steady state. Although the solution in the lower compartment could not be sampled while the steady state was being achieved, its concentration could be calculated for each sampling with a knowledge of the original concentration, the concentration in the upper compartment, and the volumes of each compartment. When a steady state transfer was established both compartments were emptied and the concentrations of materials were deter— mined directly. Fig. 11A illustrates the establishment of a constant diffusion rate for the standard KCl solution. The KCl concentrations were determined by titration with standard AgNO3 using fluorescein as an indicator. A final diffusion function of 1.62 x 10—6 sec_1 (with a maximum error of 2%) was obtained from this curve and a cell constant of 0.109 cm_2 (i 2%) was calculated by the use of equation (13). Fig. 11B shows the establishment of a constant diffusion rate for glucose dehydrogenase in 0.05 molar, pH 6.5 imidazole buffer, and Fig. 11C shows the same for the enzyme in buffer plus 5 molar NaCl. Enzyme activities rather than true concentrations were used in calculating the diffusion functions. The slopes of the curves are inverted because the membrane was initially filled with solvent 39 A. K cu STANDARD 2 3 TIME IN HOURS 8. LOW IONIC STRENGTH o o—G'Q—p- / I P’0 l I I l L 4 8 I2 I6 20 TIM E IN DAYS c. HIGH IONIC /0 3 - O \ STRENGTH / o 2 - N ,0 ,0 O l- TEMPERATURE CHANGE o I l l I I 4 8 |2 l6 20 TIME IN DAYS Figure 11. The establishment of a constant diffusion rate in the Stokes cell for 0.1 molar potassium ‘ chloride and for glucose dehydrogenase in a low and high ionic strength solvent. The low ionic strength solvent is 0.05 molar, pH 6.5 imidazole chloride buffer and the high ionic strength solvent contains 5 molar NaCl in addition to buffer. 40 rather than protein solution. Fig. 11B indicates a diffusion function of 2.1 x 10_7 (7% maximum error). This corresponds to a diffusion coefficient D of 1.9 x 10‘6 cm2/sec (i 9%) for the enzyme in a low ionic strength solvent. Judging from the curve in Fig. 11C, a steady state gradient was initially established and then destroyed by a temperature increase caused by a faulty thermo-regulator. However, with continued diffusion, the original rate was eventually obtained. The final diffusion function was 1.6 x 10-7 Q: 20%) which gives 6 cmz/sec L1 22%) a diffusion coefficient of 1.5 x 10- for the enzyme in a high ionic strength solvent. Not all of the material present in the cell initially could be accounted for at the end of the diffusion experiments. Thus, only 93% of the KCl, 85% of the enzyme in low ionic strength solvent and 77% of the enzyme in high ionic strength solvent was accounted for. These losses are too large to be attributed to non—recoverable enzyme in the membrane. They could not be due simply to denaturation either because a control solution of enzyme in the low ionic strength solvent was held for 9 days at 20 C with no detectable loss in activity. However, the porous membrane offers a very large surface for the adsorption or the denaturation of protein. The adsorptive capacity of the membrane was checked by measuring ”excesses -'.:S.CIi (3‘38 '1’.) “\E‘ -. :LIEJ‘TL; ._ mmi enigma flambo- unmn also]: I91 1 11.5.. "' \:_.I‘«| . 1:131 esw 311915539 93533 26593: a 911 .91! at man: ""ns'J-Rfl-I- 93'1'5-‘3'159'593 5 V.” E'fi'in'menb nod: has 5931313!!!“qu :.-.: " - \'.._'- i'..-~.'CI- .1 03.1%}: :11. --oar * .;:J _-.:I'1Hr.'1 b 2:!" bonus: 1 _:__T___ :3.- tir..."-*."_- - r'... -.- J .fin' ._ . .-. .-:‘.':.- «501311115 -.) , . . -' . c:'..:rmui ... l. 4,-3'511I; . LII-‘3 l I f' 41 the decrease in activity of the enzyme in the high ionic strength solution after flowing through the clean porous membrane. The first 1.1 ml lost 25% of the original activity, the next 1.8 ml lost only 2% and the rest of the solution was unaffected. Thus it appears that some adsorption does occur but that the membrane soon becomes saturated. This should not affect the final value of D unless adsoprtion of proteins changes the mechanism of transport across the membrane to something other than pure diffusion. The diffusion constant alone does not provide enough information for the calculation of a molecular weight but it does provide information about the frictional properties of particles in a given solvent according to the following equation derived by Einstein and Smoluchowski: RT D _ Nf (15) Here R is the gas constant, N is the Avogadro number, and f is the friction coefficient. The friction coefficient also appears in the Svedberg sedimentation equation: _ .1flil:iflll_ (l6) 5 " Nf where s is the sedimentation coefficient, M is the anhydrous molecular weight, p is the solution density, and V is the partial specific volume of the protein, defined as the 42 increase in volume (in ml) upon adding a gram of dry protein to an infinitely large volume of solvent. Since it is generally assumed that the frictional forces encountered during diffusion are equivalent to those encountered during sedimentation, equations (15) and (16) can be solved for f and combined to give: RTs D(l—‘7p) (17) The sedimentation coefficient is obtained from the rate of movement of the protein in a centrifugal field as follows: dx _ 2 dt — SQJX (18) where x is the radial distance from the center of rotation at time t, and 0) is the angular velocity of the rotor. A more useful equation can be obtained by integrating equation (18): 2 x = ersuat or (19) 2 ln x = ln x0 + shat or (20) s = —24%9§— log X (21) CD t 0 where x0 is the initial distance of the particle from the axis of rotation. In practice, the rate of movement of a single 4+1 -.-m'n.' c v 9911']: 2:1:1 :ru'J bmuiaas 1:.- "11. __ i far. r.: Eanjiil’» :_-n.i.:mb . I L- . ; . .- -(-.!.'! '..-ifJa 43 particle is not followed but rather the rate of movement of the boundary between the sedimenting protein solution and the pure solvent centripetal to it. Thus the position of the solution meniscus is x0 and X is the position of the boundary. The method of Hbgeboom and Kuff (1954) was used to determine the sedimentation coefficient of the glucose dehydro— genase. These authors employed a sucrose density gradient to stabilize the protein-solvent boundary formed during centrifugation and determined the location of the boundary by direct sampling of the centrifuge tube contents. In the experiment to be described a model SW—39 swinging bucket rotor was used in a model L preparative ultracentrifuge (Beckman Instruments, Spinco Div.). The rotor was equipped with a heat radiating collar into which a mercury thermometer was placed during the run. Linear sucrose density grandients were formed in 6 ml Lusteroid centrifuge tubes with the aid of the double vessel mixing device shown diagrammatically in Fig. 12. For small volumes, this device is most conveniently constructed from a solid block of Lucite by drilling out the vessels and channels. The concentration gradient was obtained by a progressive dilution of the more concentrated solution in vessel A by the less concentrated solution in vessel B as .‘ljziatumtod on: to mum on: I1 p I. .8 .1 m I . . I‘ . .-'_ .'-_ . . ' ' : I '3 . .. on: 10 nut-q ‘ - )1 -r_. 2‘ ”yr-s»? aaoartlp ed: 30 inatnfrnaoa 11013513311608 Ill: Muirfin ' -. cz' Esau saw (#391) not Dan M in m ‘ .. '... ‘ ‘l ‘ '”!'TMA Leibs: {31.21195 920-1911.: s. Pimp-._Lqmn 310(1qu 980!!! at.” tfiilflb 5rLu=z gznrnnc' fifittlr:-cifl nag 9d: ssiijdsja 09 ‘1"” “. T :n.+n:nf ‘ : Erni'z':a5 fr: noifisynfijainaa - .aifiafr‘r fi"“ t . - T"‘ ' 1‘ “F‘.1W’a j’v'ih Yd : r , - ' ." 1 ' =' "i fitnicnqxs -_~ .' -_.' . . . . '. . .. . 1'...” ...,....I.r,.i ...... 44 the system was emptied into the centrifuge tube. Rapid mixing in vessel A insured uniform dilution. Since the vessels were of equal dimensions, the rate of flow from the system was twice the rate of flow from B to A. This satisfied the requirements for a linear change in the concentration of the solution in vessel A, according to equations derived by Lakshmanan and Lieberman (1954). The gradient produced by this device was examined experimentally by adding methylene blue to the solution in vessel A. The resulting color gradient was sampled at intervals throughout the tube and the color intensity in the samples was measured at 600 mu in a Spectronic 20 spectro- photometer (Bausch and Lomb). Samples were removed from the centrifuge tube with the apparatus pictured in Fig. 13. This is essentially the same apparatus used by Hogeboom and Kuff (1954). To avoid contaminating dilute fractions with those of higher concentration, the tube was sampled from top to bottom by displacing the liquid column very slowly with a dense, 65% sucrose solution. Portions of the gradient column were completely removed and their volume measured with a graduated pipette as they flowed through the perforated plate in the sampling cup. Fig. l4 is a plot of optical density at 600 mu versus the volume removed from the tube containing the methylene Figure 12. Figure 13. 45 B A L _.J HL—J 1 a1 ’42:. “1 Density gradient forming device. /SAMPLING CUP E ’CENTRIFUGE TUBE fl, (1%.—65% sucaose Centrifuge tube sampling device. 46 blue gradient. The gradient is obviously linear. The procedure for obtaining sedimentation coefficients for glucose dehydrogenase in low and high ionic strength solvents was as follows: A solution of enzyme in 0.05 molar, pH 6.5 imidazole buffer and a similar solution containing 3% sucrose were mixed in the gradient device to produce a solution with a linear sucrose gradient and a constant concentration of enzyme and buffer. This procedure was also used for the enzyme in a buffer plus 5 molar NaCl solution. Each tube was subjected to a centrifugal field for a known time and sampled in the manner already described. The time and speed during acceleration and deceleration were recorded and expressed in terms of an equivalent time at top speed. This time (7 min in each case) was added to the total time for each run. The results of the samplings are shown in Fig. 15 as a plot of enzyme activity versus distance from the meniscus. Distances were calculated from the known volume of the fluid column and from the dimensions of the bucket, tube, and rotor (Hogeboom and Kuff, 1954). The positions of the boundaries were obtained from the inflection points of the curves in Fig. 15 and used to calculate sedimentation coefficients according to equation (21). From the sedimentation 47 \03- .. 1 E . C) 00.6' 0 <9 i—o.4- ' < s 60.2- o' . o l 1 L 1 I L o I 2 3 4 5 '6 . ML. OF SOLUTION Figure 14. A determination of the precision of the gradient forming and tube sampling devices used in the sedimentation experiments. Optical density at 600 mu reflects the methylene blue concentration in each sample. 500- 2 l [V >- ” 1' '_ 2 00 " 2 ’— L) ‘< IOO - I; High Ionic >- Strength Low Ionic ; Strength U 0 1 1 0 l 2 3 4 DISTANCE FROM MENISCUS —CM. Figure l5. Sedimentation curves for glucose dehydrogenase in low and high ionic strength solvent. The low ionic strength solvent is 0.05 molar, pH 6.5 imidazole chloride buffer and the high ionic strength solvent contains 5 molar NaCl in addition to buffer. 48 coefficients and the diffusion coefficients, the molecular weights of the enzyme in each solvent were calculated according to equation (17). These data are summarized in Table l. The viscosities shown in Table l were obtained with an Ostwald viscometer using water as a standard. In this way the characteristics of the viscometer cancel out and solution viscosities can be calculated from the following equation: n =— -—L ~nH20 (22) where t is the flow time in the viscometer and p is the density of the solution. The subscript H20 refers to the water standard. The partial specific volumes were not actually measured but were assumed on the basis of values obtained for most other proteins. The temperature of the rotor was assumed to be the same as the temperature of the thermometer. The thermometer temperature was 20 C during most of the run with brief fluctuations of i 0.2 C. To obtain D20w the following relationship of the friction coefficient to viscosity was employed: Jriw benisddo even 1 slflnT at nwoda asifiiaoollv ad! :11 .r.'_r;f J."..".ZI -'2 .- ‘..‘a O t ‘.l ‘ 1“ fi‘. ”ism rainy 'iojamnaiv bin!” 1" ' r~.=_.-:a.iusjssusdn at” Ema Hui .1 iucoaiv .1- :.'-..".. 9' I 'I. l : t n '..‘ .u'CW ' ’ 'I(.' Q . l- ‘ Table l. A summary of data used for calculating the molecular weight of glucose dehydrogenase in low and high ionic strength solvents. Low Ionic High Ionic Strength Strength (T/2 = .04) (T/Z = 5.04) Distance from axis of rotation to meniscus (x0) in cm. 5.4 i 1% 5.4 i 1% Distance from axis of rotation to boundary (x) in cm. 7.9 i 4% 5.8 i 4% Time at top speed (t) in sec. 3.40 x 104 3.67 x 104 Angular velocity (a» in radians/sec. 4.12 x 103 4.12 x 103 Solvent viscosity (n) in poise at 20 c. 1.02 x 10'2 1.87 x 10‘2 Solvent viscosity with 3% sucrose added (us) in poise at 20 c. 1.09 x 10‘2 2.06 x 10'2 Solvent density (p) in g/cm3 at 20 c. 1.00 1.19 Solvent density with 3% sucrose added (ps) in g/cm3 at 20 c. 1.01 1.20 Partial specific volume (V) in ml/g. 0.72 i 4% 0.72 i 4% Absolute temperature (T) in K. 293 293 Diffusion coefficient in 1.9 x 10‘6 1.5 x 10‘6 solvent (D) in cmz/sec. i 9% i 22% Sedimentation coefficient in 6.6 x 10‘13 1.2 x 10'13 solvent (8) in sec" . i 5% i 5% Diffusion constant in water 1.9 x 10'6 2.8 x 10'6 at 20 C (DZOW) in cm /sec. i 9% .i 22% Sedimentation constant in 6.9 x 10‘13 4.2 x 10‘13 water at 20 C (520w) in sec‘1 i 5% i 5% Anhydrous molecular weight (M) in g. 32,000 i 8,000 (max error) 13,000 i 5.000 (max error) - -.—-. ' ' ' ' "11:1: ':.i:'I '-:-' "r _ n ‘ . ' '1 Int [-.J .l- . , .J r ‘ I “-1- I ' .1 In." t..- I . II 3' :0. n ' ' I'_). .l 50 f = K n (23) where K is a friction coefficient characteristic of the particle alone and n is the solvent viscosity. By combining equation (23) with equation (15), the diffusion coefficient can be related to the solvent viscosity. If one divides this expression for diffusion in water at 20 C by a similar expression for diffusion in a particular solvent at 20 C most of the terms cancel out and the standard form of diffusion coefficient (or, actually, diffusion constant D20w) can be obtained: DZOw/D = n/nzow or (24) D20w = ——T)— . D (25) T‘20w The same type of calculation was used to obtain 820w from equation (16). The average viscosity over the distance moved by each boundary was used in these calculations. It should be noted that the per cent errors included in Table l are maximal errors based on the maximal deviations from the best estimate of the data in Figs. 11 and 15, and on the range of values given for the partial specific volumes of various proteins by Fox and Foster (1957). In some cases, the probability of the maximal error may be very low. For example, in Fig. llB, considerably more confidence 1 .J__:- :_ 9151's}: ‘ ‘1. ‘ ‘- _: . .2-" ' "IV-n - If. II o ‘ \ I . / I . 51 can be placed in the value selected for the diffusion function because more data in the steady state region were obtained. Errors of less than 1% were ignored in the calculations. It is readily apparent that the glucose dehydrogenase dissociates in high ionic strength into subunits. One would expect the dissociation to yield subunits of 1/2 or 1/4 rather than 12/32 the weight of the molecule in low ionic strength solvent. The unusual ratio of molecular weights observed could result from the high degree of error in these calculations, especially in the assumption that the partial specific volume is not affected by changes in ionic strength. If the value of 9 in high ionic strength solvent was lower than the assumed value, the actual value of M could be some- what lower than that given in Table 1. It is also possible that the calculated value of M in 5 molar NaCl is a weight average value for a mixture of the low and high molecular weight units. This too would lead to an erroneously high value of M for the subunit. In spite of the errors involved in the calculation of molecular weights, it is apparent that a depolymerization rather than a polymerization occurs as the ionic strength of the protein solution is increased. 52 Since it was obvious that the increase in thermal stability in a high ionic strength solvent was not the result of intermolecular polymerization, the possibility of additional intramolecular hydrogen bonding was considered. Guanidine is strongly hydrogen bonded in aqueous solution and, therefore, is capable of rupturing existent hydrogen bonds and forming new ones. This is thought 03 be the reason for its inactivation of proteins. If increased ionic strength enhances heat resistance by strengthening hydrogen bonds or forming new bonds, it should protect the enzyme from guanidine inactivation. The inactivation rate of the enzyme can be expressed by the following equation: —d(E)/dt = a(G)n (26) where the change in enzyme concentration with respect to time is equal to a proportionality constant times the con— centration of guanidine raised to a pOWer. The reciprocal of the time required for a given fractional inactivation of the enzyme is also an expression for the inactivation rate: -d(E)/dt = F(l/t (27) 40%) In the present study the time for 40% inactivation was used. Then: 1/ a(G)n (28) t40% saw szv .' of: :'.i. --r.- .. :e;c:c_. and ,m iiiZSLIAtrnginq . '.-..-.r :. '. _-'.-..--' I:‘-.'_f-'.ib‘ ii _.s"xir:t:-I.nm513n.i :2- .-: . .! . '. -.- . . ' r: -- .". '- .-..g_-nu.1:1-- a}. ‘bfls -_.-;'_' 1.1. 1,‘ .--‘.'-5.::;nn3 i. _-'..'--'.. " 1:. ..f.r:r»d. 53 By taking the logarithm of each side the equation becomes: log l/t4O = log a + nlog(G) (29) When plotting log l/t40 versus log (G) the slope of the line indicates the number of moles of guanidine needed to inacti— vate one mole of enzyme. The procedure for the guanidine inactivations was as follows: Samples of purified glucose dehydrogenase in 0.05 molar, pH 6.5 imidazole buffer and samples of enzyme in buffer plus 5 molar NaCl were dilutedJJl with an appropriate concentration of pH 6.5 guanidine. The inactivations were run at O C in crushed ice. At various time intervals, 0.1 ml of each enzyme—guanidine mixture was removed, diluted 9-fold immediately, and assayed. Fig. 16 and 17 are a series of progress curves for the inactivation of glucose dehydrogenase under the conditions described. The numbers at the end of each curve indicate the final concentration of guanidine in the inacti— vation mixture. It is obvious from these curves that the 2.5 molar NaCl concentration affords some protection for the enzyme from guanidine inactivation. It is also apparent that, even with no salt present, the guanidine inactivation of this enzyme is remarkably slow. .®>.H.DU SUMO MO UCQ 03» um um¢wuapnfl mum mGOHpmnquosoo wcflpflcmsw .U 0 pm uwmmsn wHonoflEH m.m mm .HMHOE mNo.o QH wmmcwmoupmsmp mmOUSHm mo goaum>HDUMGH wcflpflgmsw .wa musmflm 9.1301 2. m2; mm \ .. ON 0. o_ m o _ _ .-. u _ O 0 ON 4 5 o \ O 2o.wI.Q| III IIIIIIIIIII IIIIo . ov o b O I .oo \ o 2 m.N .0 O \ \ O I oo 0 20 m\ \ 2m m O\o I oo— 1N3383d NOIiVAllDVNI .9630 some mo one 93 um amt/Hm mum mgonmumupnmocoo onflwanmsw .0 0 pm Homz Hmaofi m.N mafia memsn wHONmUHEH m.m mm .HmHOE mmo.o nun mmmcwmouwhflwp mmoosam mo Goflpm>flpowgfl wcflpflcmsw .ha musmflm Wyn—DOT. 2. NE; on O? on ON 0. J 1N3383d NOIiVAILQVN I 56 Fig. 18 is a plot of log l/t versus log of the 40% guanidine concentration according to equation (29). The slopes are the same for the enzyme in both low and high ionic strength medium suggesting that no additional hydrogen bonds are formed due to the high salt concentration; or at least no bonds affecting enzyme activity are formed. The slopes are 8.3 indicating an 8th order reaction with respect to guanidine. In other words, 4 hydrogen bonds must be broken to inactivate one molecule of enzyme assuming that two molecules of guanidine are required to break a hydrogen bond. In interpreting these data it is also assumed that guanidine disrupts only hydrogen bonds. In reality this may not be the case since there has been some evidence that hydrophobic bonds are also weakened by guanidine (Kauzmann, 1959). 0.0 1.40% Figure 18. L06 OF ‘ l C) 57 Low IONIC STRENGTH— HIGH IONIC o _STRENGTH I I I I O I I I I I I I I l 0 i ._ 0 -___I SLOPE:8.3 ’0 I I I l _ 0.2 0.4 ' 0.6 0.8 LOG OF GUANIDINE CONCENTRATION The effect of guanidine concentration on the inactivation of glucose dehydrogenase in a low and high ionic strength solvent at O C. Solvent conditions are given in Figs. 16 and 17. DISCUSSION The purpose of the research described in this thesis was the isolation of a spore enzyme which could serve as a model for studying spore heat resistance. Information on the factors affecting the heat resistance of this model could be used to formulate a mechanism for the heat resistance of intact spores. Thus the use of a model would permit the investigation of spore heat resistance without the uncertainties inherent in viability measurements. Glucose dehydrogenase is a spore enzyme (Bach, 1961) and is similar to whole spores in some of its properties. For example, this enzyme is extremely heat stable in intact dormant spores, which are themselves heat stable, and relatively heat labile in intact germinated spores. Under the proper conditions of pH and ionic strength, the purified enzyme is as stable in yitrg as in the intact spores. Thus the glucose dehydrogenase is a very suitable spore model and should prove to be a powerful tool for studying the mechanism of spore heat resistance. The possibility that spore macromolecules are protected from thermal denaturation by a highly ionic spore interior 58 I _I. -,1-'.1.f 31.! :--.-xI'-'-1:..-e:b Lawson-J mi:- 30 5801111."; MT T-..-.u- Ij--.I.I .- -.r_-r.-.-:.=-v _._..-;-., r. '.Iu nuiJIsloai M3 In =:!..' ._ . ' If“? 1'11; ' '. 'I.I ' .'.. _-.';.' '__!u 5?- 10:. 19M --._.-. :--. 1.3951 9d: , f. . \ -. .J I-:‘ bow.- at; ) I . I.'i.i.' . I n. -VL l I. ..I I. U. - I \ I I -.. V. H‘- \ x 59 is very intriguing because it agrees with much of the pre— vailing data on spore properties. For example, spores accumulate high concentrations of calcium and DPA and thermal stability is directly related to the concentration of these compounds in spores. Also, the loss in spore heat stability during germination is accompanied by a loss in calcium and DPA. The fact that spores are permeable to water and other solutes refutes the hypothesis of a completely anhydrous and therefore heat stable spore, but supports the increased ionic strength concept of stabilization. Although the catalase and glucose—6—phosphate dehydrogenase from spores were only slightly protected from thermal inactivation by increased ionic strength, no final conclusions about their stabilities should be drawn until the effect of pH on these enzymes is studied. Protein structure can be treated at three different levels of complexity: the primary structure or the sequence of amino acids in the peptide chain; the secondary structure characterized by a regular coiling of the peptide chain into a rigid helix; and the tertiary structure or the folding of the helix into the final three dimensional configuration of the protein molecule. The specificity of the tertiary structure is thought to be responsible for the specificity '."s': ' 5 mm. 1171:! -.:.-___:I.':rr1_" b 2.2 gift 1 . -.-'.;I .I- in .4109: I11 abrmoqmos i. u"?:'."--'.'...Ifl'19[‘ pnfnwb ;.'- 'III'I.‘ .A‘II'I hm: .1 - '-;'.I:'(:. 15.23:- s -. -' 'ri "t‘:.""1.l‘-|.'..: brr'z .. '1.’; I £35. ' '.I. ”H": ‘ .. . i - 60 of enzyme catalysis and, conversely, the loss of this structure is thought to be responsible for the thermal inactivation of enzymes. If this is true, the effect of various solvents on the thermal stability of glucose dehydrogenase is probably due to effects on the forces which maintain tertiary structure. These forces should therefore be considered in discussing a possible mechanism for the effect of pH and ionic strength on the stability of glucose dehydrogenase. There are ample opportunities for hydrogen bonding between various groups in proteins and this type of bond probably plays a very important role in maintaining protein configuration at both the secondary and tertiary levels. At the secondary level, hydrogen bonding occurs principally between amide nitrogens and carbonyl oxygens of the peptide backbone while, at the tertiary level, the side chain groups of amino acids are involved. If it is assumed for the moment that the thermal inactivation of glucose dehydrogenase results from the rupture of hydrogen bonds, the probable effect of pH and ionic strength on these bonds, and hence on enzyme activity, can be considered. According to Scheraga (1961) the strength of hydrogen bonds between ionizable amino acid side chain groups is dependent on the state of ionization of the groups and thus on the pH of the solution. If bonding occurs between two .15 811:.- F ".. .Il Tm 15:1 'r I. i.-'ir:t:.-. ifimndi SH .33!) 03 sub '17. um 'Ic MIT '- r415 -.,L:i.." \.'.i'(.-q 61 similar groups such as the imid nitrogens of histidine: \ + 2 N_H no-ooN / \ the bond would be strongest when half of the imidazolyl groups of the histidine residues were ionized. In view of this conclusion, it is of interest that the pK of the imidazolyl group (6.00) is close to the pH of maximum stability (pH 6.5) of glucose dehydrogenase. However, it is difficult to believe that histidine could be involved because the inactivations were performed in an imidazole buffer. The concentration of imidazole in this solution was 5 x 10_2 molar and the concentration of enzyme was only 3 x 10_8 molar. Thus, one would expect competitive inhibition of any histidine-histidine hydrogen bonds by the imidazole. On the other hand, the pH dependent maximum in stability may result from a combination of the ionization characteristics of several heterologous bonding groups. This might explain the difference in the slope (reaction order) on either side of the minimum of log k seen in Fig. 6. For example, in a cooperative hydrogen bond between a carboxyl acceptor group and a pair of amino or phenolic donor groups, two ionizations can occur at high pH and only one at low pH as follows: /0'”-'HO-C < OH— //0--H—0—c < 0H //0 0-c < -c -C —c \ ‘ + \ _ _ \ 0H HO—C < H+ 07-H—0—c < H 0 0-c < ‘1. '.--._r. Imriv. J-.')"_-u()‘..3&’. 9d biuow band Fran. .' .mf :'E‘.'JI finibiiald ad: 10 - . - . , If . I qmnp ., . -. ' ' " in .- -. r .1“, . , .f. v- r. I . \ 62 The stabilizing effect of a high NaCl concentration can also be explained in terms of hydrogen bonding. Huggins (1962) discussed the strength of hydrogen bonds of the type: " / _?1H....o=c\ in terms of the relative distance of the hydrogen from the two bridge-head atoms, nitrogen and oxygen. It appears that the more symmetrical the bond, the stronger it is. If the amide nitrogen is made more electropositive by binding a proton or metal ion, the hydrogen is displaced toward the oxygen, increas— ing the symmetry and the strength of the bond. The binding of Na ions by the amide nitrogens in proteins should also decrease the rate of inactivation by guanidine, unless the competing protein—guanidine hydrogen bonds are strengthened to the same degree as the protein—protein hydrogen bonds. The electrostatic bond or salt linkage has also been suggested as a factor in the maintenance of protein configur— ation. The variation in the stability of glucose dehydro- genase with varying pH could be due to salt linkages. It seems unlikely, though, that bonding of this nature is responsible for the stability of the enzyme in high ionic strength solutions because high concentrations of ionic material tend to rupture salt linkages by forming ion clouds around the bonding charges. On the other hand, the rupture '..I-.'. ' " .'il. dfiguhuflu ad: fiestas-lb :..'.'r I u o .1_--;':- J.ir av nIiw: ed: lo smug: at -. . ,-.I:-. 1r. hrmd—z-gblld '. '-:""Ir,-II " 'r ”n on: r - l' u :3.- .. ‘.. \ ' \, Jr'dl | ‘ ' .71.. I. K \ 63 of salt linkages between peptide chains could be causing the dissociation of the glucose dehydrogenase molecule in high ionic strength solution. If glucose dehydrogenase is folded with polarizable and ionizable groups on the surface and with non—polar groups oriented toward the center of the molecule, a competition between extension of the molecule through solvation and con- traction of the molecule through hydrophobic bonding would occur. Thus, a decrease in the solvation of the protein should strengthen the existent hydrophobic bonds. A decrease in solvation could occur at some particular pH when the surface charge was at a minimum or perhaps at a favorable balance. It could also occur in high ionic strength solvents because the dense ion clouds surrounding the protein charges would prevent the binding of water molecules. It is of interest in this connection that high concentrations of guanidine may rupture hydrophobic bonds as well as hydrogen bonds (Kauzmann, 1959) especially at low temperatures. It should be noted, also, that the strength of hydrophobic bonds increases, within limits, with increasing temperature. The rupture of protein hydrogen bonds in aqueous solution probably occurs through an exchange reaction with the hydrogen bonds between water molecules. Thus the . I. - ' . . "fill-3:. I'I .. .I'Ij .nojjuloa dipiszil .:q 1.11 bob -3 ai ocangnubydab annoulp II L-~ rf‘w bun ,rnhvna Id1 no aqu019 oldsstnoj baa -nr 5 (n1 :u“hm 1r: .. 1 :u u '35 I‘Luoj b53n9110 . . - . ‘ - . = _.- . . - :.. .LiI-"Z’ r . I.Ia-.in-:>LZ .cm I .- H-Iflwnri 'l "'rI . . .— ll . l . ,,. '.I'..I . .- D . 1 I 64 effect of high salt concentration on enzyme denaturation could be caused by a lowered water activity as well as by direct effects on the strength of bonds in the protein. The very limited effect of concentrated sucrose solutions on the rate of inactivation argues against the idea that lowered water activity is the basis of thermal resistance. However, it should be pointed out that the lowering of water activity by salts is not only due to dilution of the water but is also a consequence of the binding of polarized water molecules by charged ions (Brey, 1958). The protection of proteins from thermal denaturation by high salt concentrations has been observed by other workers. Harrington and Schellman (1957) (as reviewed by Shooter, 1960) studied the configuration of ribonuclease in solution by a combination of optical rotation, viscosity and UV absorption measurements. They found that high concentrations of LiBr (up to 9.9 molar) prevented the configurational changes observed when this enzyme was heated in water alone. These authors concluded that the addition of LiBr to the solution increases the amount of helical structure in both native and oxidized ribonuclease. The use of polarimetry ‘or some other non-enzymic method of measuring protein denaturation could be of value uf ”WHO. :I nignwrse -nr no 3399119 . ” ..h12na hojsxdnnanos 10 35931. 5.11:1! VII! .1! .nri ufij janir;s noun!“ nai‘rviflofini lo .301 on: \ ' .'... .5 ..'..-2."? .- ' ‘_'-'-. 2 _ r ". " '."- . I'i E'. i. ‘1'“ '11:!!! IOSIW ‘. . - . :- ' - - .' "I {111118.11 A 65 in future studies of the glucose dehydrogenase because it would permit the direct measurement of configurational changes in any solvent and at any pH. With the present enzymic method, only irreversible changes in configuration can be measured unless the changes occur near pH 8. Of course, it may be difficult to correlate the configurational changes with changes in enzymic activity. In any case, the use of physical or chemical methods for examining configur— ational changes is going to require large quantities of pure glucose dehydrogenase. Perhaps one of the more important contributions of the present study is the information on molecular size and conditions of maximum stability for this protein, since this information may be used to design a more efficient purification procedure. J: has shsvioa 2n: icinvoiri 'lao ,bodjll :r.- .. "I : "- E’Ie‘lf1.'.:.. -'.-I‘.: asemu tmtumanl ad use '3r _f“ _.: n:_ . I“ : .-'r“fii on 35m 3i ,aazuag ' :r z. . '1... " H: - 131w aswnsde gnu :u sun SUMMARY Glucose dehydrogenase from Bacillus cereus was examined as a possible enzymic model for the study of spore heat resistance. In 2129 studies showed this enzyme to be relatively labile in intact sporulating cells and germinated spores but very stable in intact dormant spores. The heat resistance of both the glucose dehydrogenase and a glucose— 6—phosphate dehydrogenase in spore extracts was dependent on the concentration of dialysable material in the extract. Glucose dehydrogenase from germinated spores was purified 3,400 fold and the effect of various solvents on the properties of this enzyme were investigated in an effort to understand the mechanism of its heat resistance in spores. Heat resistance of this enzyme is strongly dependent on hydrogen ion concentration,exhibiting a sharp peak at pH 6.5 in .05 molar imidazole buffer. The Arrhenius activation constant for this enzyme in the same buffer is 17,000 K_l. High concentrations of certain ionized compounds such as NaCl, KCl, Na2S04, (NH4)2SO4, DPA, and calcium—DPA chelate also increase the enzyme heat stability. The increase in stability in NaCl is approximately 2nd order with respect to the concentration of NaCl. The effect of CaCl2 on heat 66 .1e flu. 2'? 'w on siquus oidjaaoq n In boat-Ina .II . . t .'.l l . .. . _- ; 51.: -.: IN- ..1..:-.v.r. 3519:. . - " ‘.. ..' .'.' ' I. -.'. ' rig??- _ I I ,J- r «'1'. l I. I II ' :—( . ..' . . . \ , , .. -- . .. .". :.:.. r-_...-. n_£_ .-_-.-:n'51a1u91. 3mm 67 resistance was also examined and found to be more complex than the NaCl effect. At concentrations up to 0.8 molar, CaCl2 protects the enzyme from heat denaturation while, at con— centrations above 0.8 molar CaClZ, the enzyme was rapidly inactivated. Purified glucose dehydrogenase is very resistant to guanidine inactivation at 0 C and the presence of 2.5 molar NaCl increases this resistance. The inactivation reaction is 8th order with respect to guanidine concentration and does not change in the presence of the NaCl. The molecular weight of purified glucose dehydrogenase in a low and high ionic strength solvent was determined by a combination of diffusion and sedimentation measurements. In 0.05 molar, pH 6.5 imidazole buffer, the molecular weight was approximately 32,000 while in buffer plus 5 molar NaCl the molecular weight was approximately 13,000. Although it is entirely possible that a highly ionic environment is responsible for the heat resistance of this enzyme in intact spores, the evidence obtained to date is not sufficient to prove this thesis. The mechanism of the effect of pH and ionic strength on the heat stability of purified glucose dehydrogenase also cannot be stated with certainty. It appears to involve the strengthening of side— chain hydrogen bonds and possibly hydrophobic bonds responsible aiij-1u35nou issJImvir )mvsns 9d: -: um 1n= 9d: .cLDuD 151cm 8.0 svods anoijszdnoo .bnssvisasnl " raf-- _C’ I II _ . _ E, j -‘I'J.: :n:..: 1; aIibtnsuL -. ‘._ ' ..‘.i-I h \ - \ - \ ‘- 68 for maintaining the native configuration of this protein. LITE RATURE CI TED Bach, J. A. 1961. The glucose dehydrogenase of Bacillus cereus; a model for the study of spore heat resistance. M.S. Thesis, Michigan State University, East Lansing. Bach, J. A. and H. L. Sadoff. 1962. Aerobic sporulating bacteria I. Glucose dehydrogenase of Bacillus cereus. J. Bacteriol. 83:699—707. Black, S. H., T. Hashimoto, and P. Gerhardt. 1960. Calcium reversal of the heat susceptibility and dipicolinate deficiency of spores formed Vendotrophically? in water. Can. J. Microbiol. 6:213-224. Black, S. H. and P. Gerhardt. 1962. Permeability of bacterial spores. IV. Water content, uptake, and distribution. J. Bacteriol. 83:960-967. Brey, W. 8., Jr. 1958. Principles of Physical Chemistry, pp. 159—161. New York: Appleton—Century—Crofts, Inc. Curran’, H. R., B. C. Brunstetter, and A. T. Myers. 1943. Spectrochemical analysis of vegetative cells and spores of bacteria. J. Bacteriol. 45:484—494. Doi, R., H. Halvorson, and B. D. Church. 1959. Intermediate metabolism of aerobic spores. III. The mechanism of glucose and hexose phosphate oxidation in extracts of Bacillus cereus spores. J. Bacteriol. 77:43-53. Dyrmont, A. 1886. Einige Beobachtungen uber die Miltzbrand- bacillin. Arch. of Exptl. Pathol. Pharmakol. 21: 309—317. Fox, W. F. and J. F. Foster. 1957. Introduction to protein chemistry. New York: John Wiley and Sons, Inc., p. 218. Friedman, C. A. and B. S. Henry. 1938. Bound water content of vegetative and spore forms of bacteria. J. Bacteriol. 36:99—105. Gerhardt, P. and S. H. Black. 1961. Permeability of bacterial spores. II. Molecular factors affecting solute permeation. J. Bacteriol. 82:750-760. 69 . .rl J‘I - - .. .... . - rh- ' |. ..L.-'-‘I...A. IA .- 1 _u . 1b.3- ouJ ..3 rabom s :532122 \- i.i~vinb 535:3 nrgiduin .ILIOHT .I.l . . I, I. : ... .l -. 2‘ 'Jl'. u S'- f.‘ I. I .la(-bfi?l II] OH “I 0‘ 0‘. .‘m ;.3.uh ._ : run.m;hgnui :ncou'D .I simsto‘d .. I 'T A: .IDJITJSLfl .L 7O Goldberg, C. A. J. 1959. A discontinuous buffer system for paper electrophoresis of human haemoglobin. Clin. Chem. 5:446. Gortner. R. A. 1930. The state of water in colloidal and living systems. Trans. Faraday Soc. 26:678—686. Halvorson, H. and C. waitt. 1961. The role of DPA in bacterial spores. In H. O. Halvorson (ed.) Spores II. Minneapolis: Burgess Publishing Co. Halvorson, H. O. 1957. Rapid and simultaneous sporulation. J. Appl. Bacteriol. 29:305—314. Harrington, W. F. and J. A. Schellman. 1957. The effect of concentrated solutions of lithium bromide on the configurations of polypeptides and proteins. C. R. Lab. Carlsberg, Ser. chim. 39:167—182. Henry, B. S. and C. A. Friedman. 1937. The water content of bacterial spores. J. Bacteriol. 33:323—329. Hogeboom, G. H. and E. L. Kuff. 1954. Sedimentation behavior of proteins and other materials in a horizontal preparative rotor. J. Biol. Chem. 219:733—751. Huggins, M. L. 1962. Physicochemical aspects of hydrogen bonds and their application to biology. American Scientist. 59:485-496. Kauzmann, W. 1959. Some factors in the interpretation of protein denaturation. Adv. Prot. Chem. 14:1—63. Koffler, H. 1957. Protoplasmic differences between mesophiles and thermophiles. Bacteriol. Rev. 21:227—240. Lakshmanan, T. K. and S. Lieberman. 1954. An improved method of gradient elution chromatography and its application to the separation of urinary ketosteroids. Arch. Biochem. Biophys. 53:258. Lawrence, N. and H. O. Halvorson. 1954. Studies on spores of aerobic bacilli. IV. A heat resistant catalase from spores of Bacillus terminalis. J. Bacteriol. 68:334-337. ...-"". .'- .'.LI -'- "U L ‘5 '3 A-JI' lukn;'i .anszT .amsjaga . ’.-'_-"-‘- *‘u'I-T‘ .J'.-1"[ .jjj‘s-edl .D has .F nae-SWIG}! ...r} “firizv' 1 .5 .E r; .Lsroqa lsluajsld . .:-..e.'-.-_-_:':‘v ar --.'.-.-."'- err arr-snail! .II 2970!!! . . L" . -i- " ..'2. .0 .V ,nnauovlsh . - .: . _ . ".:J' .4.r_:t:'. .i .. _ - - - - n . ‘ I311..- 'n'f'. . fl 1 - c ‘ v 9 . . u \- L o ‘ n a - - \ i v \ - a o \ \ \ b \ 71 Levy, M. and A. E. Benaglia. 1950. The influence of temperature and pH upon the rate of denaturation of ricin. J. Biol. Chem. 186:829—847. Lewis, J. C., N. S. Snell, and H. K. Burr. 1960. Water permeability of bacterial spores and the concept of the contractile cortex. Science 132:544—545. Lewis, J. C. 1961. Discussion. In H. O. Halvorson (ed.). Spores II. Minneapolis: Burgess Publishing Co. Lewith, S. 1890. Uber die Ursache der widerstandsfahigkeit der Sporengegen hohe temperaturen. Arch. Exptl. Pathol. Pharmakol. 26:341. Militzer, W., T. B. Sonderegger, L. C. Tuttle and C. E. Georgi. 1950. Thermal enzymes. II. Cytochromes. Arch. Biochem. Biophys. 26:299—306. Murrell, W. G. 1961. Discussion. In H. 0. Halvorson (ed.). Spores II. Minneapolis: Burgess Publishing Co. Powell, J. F. 1953. Isolation of dipicolinic acid (pyridine 2:6—dicarboxy1ic acid) from spores of B. megaterium. Biochem. J. 54:210—211. Powell, J. F. and J. R. Hunter. 1956. Adenosine deaminase and ribosidase in spores of Bacillus cereus. Biochem. J. (London) 62:381-387. Rode, L. J. and J. W. Foster. 1960. Mechanical germination of bacterial spores. Proc. Natl. Acad. Sci. U.S. 46:118—128. Ross, K. F. A. and E. Billing. 1957. The water and solid content of living bacterial spores and vegetative cells as indicated by refractive index measurements. J. Gen. Microbiol. 16:418-425. Scheraga, H. A. 1961. Protein structure. New York and London: Academic Press, pp. 119-125. Shooter, E. M. 1960. The configuration of proteins in solution. Prog. in Biophys. and Biophys. Chem. 10:196—236. . . ,, l. . . .-..£"i ‘.L"". .ti .=-' .'.' u.- .l'.l_-: :_».5'.tm:i:.-.-:r: "-:. yzfiiidz'fiflfllsq n n; .arjunn oLJdnqunce 9d: 10 c J I _ r n . .I_ ll . ..'I' _i_l Luvs" ILr'. - 1.0-Cl ID at! ‘BIWQJ. u . "-;r2 ..1 rfrrugi” .II ss1oqa '-'. .' "' .‘.!z-L'I. .'cl .n'J-iw-u' . . . r. ' 1 | :35. 1.‘.'-'[J . . " ' ' ‘ . ' if"? . - - \ u . \u \ .:. I - - - - \ - " . _ ‘ \ ~ I " ‘l \ . ‘ \ \ \ \ 72 Stewart, B. T. and H. O. Halvorson. 1953. Studies on the spores of aerobic bacteria. I. The occurrence of alanine racemase. J. Bacteriol. 66:160—156. Stewart, B. T. and H. 0. Halvorson. 1954. Studies on the spores of aerobic bacteria. II. The properties of an extracted heat stable enzyme. Arch. Biochem. Biophys. 62:168—178. Vinter, V. 1961. The formation of cystine-rich structure in sporulating cells and its possible role in the resistance of spores. IQ H. O. Halvorson (ed.). Spores II. Minneapolis: Burgess Publishing Co. Virtanen, A. I., and L. Pulkki. 1933. Biochemische Untersuchunger fiber Bakteriensporen. Arch. Mikrobiol. 4:99-122. Waldham, D. G. and H. O. Halvorson. 1954. Studies on the relationship between equilibrium vapor pressure and moisture content of bacterial endOSpores. Appl. Microbiol. 6:333. Warburg, O. and W. Christian. 1942. Isolierung und Kristalli— sation des Garungsferments Enolase. Biochem. Z., 310:384—421. YOung, I. E. 1959. A relationship between the free amino acid pool, dipicolinic acid, and calcium from resting spores of Bacillus megaterium. Can. J. Microbiol. 6:197-202. .r:;—:';v r..'. -' ..'-t E-m'. -‘." .El .mjaojond "idolas 10 Bazoqa; sldsjn jcad bednnijxo a: lo .ViI-fiol: i .aydqoifl .\ .: L ‘Iojniv 30 . .ruri 5 5' .’id'f .V . i :1)» r Jul'xrqa n- . --'.' . ' '-- .-;. '..r.:':*:'.' - _I‘. :' . . ‘ _rn':.. - \ . \ .' ' l 'I \ \ INLY ROOM USE INLY - WHIHINIH m«IHIMMIW“Nunlulmmml 1 2 3 5 2 8 0 3 0 3 9 2 1 3