Invertase and Its Naturally Occurring Inhibitor'

Plant Physiol. (1971) 48, 366-370

An Examination of Methods Used to Assay Potato Tuber Invertase and Its Naturally Occurring Inhibitor'

Received for publication January 18, 1971 E. E. EWING AND MARTHA H. McADO2 Department of Vegetable Crops, Cornell University, Ithaca, New York 14850

ABSTRACT inhibitor complex is actually of low dissociability, then the equations used workers Confirming an earlier report, it was shown that the en- by previous (5, 9, 16) would be in inhibitor of serious error (2, 4). Therefore, it seems worthwhile to examine dogenous potato tuber invertase forms an es. further the dissociability of the enzyme-inhibitor sentially undissociable complex with the enzyme. Conse. complex. quently, several previous analyses of potato tuber invertase MATERIALS AND which were based on equations derived for highly dissociable METHODS enzvme-inhibitor complexes are presumed to be in serious Crude extracts from potato (Solanum tuberosum L.) tubers error. The complex formation proceeded slowly, requiring were prepared for assay as described by Pressey and Shaw approximately 1 day to reach completion at 2 C, and 1 hr (14), including the use of an Acme Juicerator for homogeni- at 37 C. Allowing complex formation to reach completion zation. A shortened form of the procedure outlined by Pressey before assaying enzyme activity did not affect the noncompeti. (12) was followed for partial purification of invertase inhibitor, tive nature of the inhibition. using only the acid precipitation, fractionation with ammonium Contrary to previous reports that inhibitor could be se- sulfate, adsorption on alumina gel, and Sephadex G-100 steps. lectively inactivated through foaming in a blender, both en- Invertase purification was accomplished by the procedure of zyme and inhibitor appeared to be denatured by such treat- Pressey (12), except that crude extracts were subjected to ment. Foaming accomplished by passing nitrogen gas bubbles longer foaming periods in an attempt to inactivate the inhibitor through extracts gave more favorable results. more completely. Enzyme and inhibitor preparations were buffered with acetate at pH 4.75. Pressey's (11) definition of invertase activity was adopted whereby 1 unit is that amount of enzyme which liberates 1 ,umole of reducing sugar per hr under the conditions of the assay. Pressey's (11) assay was employed with slight modifications. The purposes of this communication are to point out con- The volume of the incubation mixture was reduced to 1 ml. flicts present in published reports of a naturally occurring The mixture was 0.1 M with respect to acetate buffer, pH 4.75, inhibitor of potato invertase; to support the view that the in- and 0.25 M with respect to sucrose. This sucrose concentration vertase-inhibitor complex is of low dissociability; to demon- gave the optimal reaction rate for the range of enzyme con- strate that the complex has an extremely slow rate of formation, centrations used in our experiments. The assay was performed especially at low temperatures; and to examine the degree to at 37 C for 1 hr. The reaction was linear with time throughout which the inhibitor is selectively inactivated by foaming. the assay period. One milliliter of copper reagent (17), which Schwimmer et al. (16) concluded that there was evidence for completely stops invertase activity, was added to terminate the the presence of an endogenous inhibitor of invertase activity reaction. Tubes were immediately heated for 20 min in a in potato tubers. Pressey (11, 12) convincingly demonstrated boiling water bath. After cooling, 1 ml of arsenomolybdate that such an inhibitor was present, reported on its partial color reagent (10) and 7 ml of water were added. Solutions purification, and characterized the inhibitor as a protein with were then mixed and centrifuged 10 min at 500g, and their a molecular weight of about 17,000. By means of Lineweaver- absorbance was read at 520 nm. Boiled enzyme blanks in- Burk plots, he showed that the inhibition was noncompetitive. cluded as controls did not vary significantly in absorbance Several workers (5, 9, 16) applied equations based on non- from reagent blanks. competitive, highly dissociable enzyme-inhibitor complex for- Foaming of solutions was performed with a VirTis blender mation to determine the amount of invertase present in tuber (12, 14) set on medium speed. The homogenizing flask was kept extracts. Pressey (11) mentioned that the reciprocal of enzyme in a water bath at 37 C during blending. An alternative method activity plotted against inhibitor concentration after the of foaming employed N2 gas bubbled into 0.5 ml of extract method of Dixon (1) gave a straight line. This would indicate through a hypodermic syringe needle, the end of which was a highly dissociable complex. However, a later paper (12) stated, cut off square. (Diameter of the needle did not appear to affect without presenting data, that the enzyme-inhibitor complex results.) All solutions foamed were 0.1 M in acetate buffer, pH showed characteristics of low dissociability (6). If the enzyme- 4.75. To compensate for evaporation during foaming, solutions were restored to their original volume by adding a suitable volume of the acetate buffer before assay. ' Supported by Hatch funds granted to Cornell University. Paper 611 of the Department of Vegetable Crops, Cornell Uni- versity. RESULTS 2Present address: 505 W. University Parkway, Baltimore, Md. From previous work (6, 7) with naturally occurring, protein- 21210. like inhibitors, it seemed possible that the combination of in- 366 Plant Physiol. Vol. 48, 1971 POTATO TUBER INVERTASE AND INHIBITOR 367 vertase with the endogenous inhibitor would proceed at a rate sufficiently slow to be measured. To test this possibility, 80 partially purified inhibitor was added to partially purified en- 6 zyme at various intervals prior to the actual invertase assay and preincubated at 2 or 37 C (Fig. 1). To check enzyme I60 stability, the enzyme was also preincubated in the absence of 4 inhibitor. With no preincubation, the amount of inhibitor used 0- A40 -2 reduced enzyme activity by 18%. Preincubation at 2 C further . reduced the activity to only 49% of the control, but consider- 2 I able preincubation time was required to reach the maximal A20 inhibition. For example, 6 hr of preincubation gave inhibition -a about half-way between that obtained without preincubation 1!... L. and the maximal level. At 37 C maximal inhibition was ob- 0 5 10 15 20 25 tained with much shorter preincubation time. Thus 1 day at Inhibitor, pi per 0.846 units invertase 2 C and 1 hr at 37 C each gave 49% of the control activity. FIG. 3. Effect of various amounts of inhibitor added to a given Presumably much of the inhibition which was noted in the quantity of enzyme. Results expressed as reciprocal of reaction absence of a preincubation period or following short preincu- rate ( ) and as percentage of inhibition (---). Vertical line at bation reflects formation of the enzyme-inhibitor complex each point denotes standard error (20 replications). which occurred during the 1-hr assay at 37 C. In spite of the longer preincubation time required at 2 than at 37 C, the lower temperature seemed preferable because the enzyme showed a 80 ~ i = 25 pi gradual loss of activity when preincubated alone at 37 C (Fig. 1). Based on these results, in subsequent experiments mixtures 60 of enzyme and inhibitor were routinely preincubated for 2 days at 2 C before assaying. The preincubation period was extended 0- to 3 or 4 days when working with extremely dilute solutions of .2_ 40 enzyme and inhibitor. The instability of enzyme at 37 C raised the question of q2nLu

I 1.00' Enzyme only 0 0.4 0.8 0.75 Invertase, units/sample 0.50 - Enzyme + inhibitor FIG. 4. Percentage of inhibition by a given quantity of in- 0 1 2 3 hibitor for varying amounts of invertase. Vertical line at each .CD E Pre-incubation time at 2', days point denotes standard error (four replications).

Enzyme only inhibitor stability under these conditions. Therefore, inhibitor solution was placed in a shaker bath at 37 C for various inter- -____!______4 vals up to 4 hr. Enzyme was then added, and the mixture was + inhibitor preincubated and assayed. It was found that even 4 hr at 37 C 3 4 had no detectable effect on inhibitor activity (data not shown). Pre-incubation time at 371, hrs Since Pressey (11) made no mention of a preincubation of FIG. 1. Effect of preincubation time and temperature on in- enzyme and inhibitor, it seemed desirable to repeat the experi- vertase reaction rate in presence and absence of inhibitor. Par- ments from which he concluded that inhibition was non- tially purified enzyme and inhibitor preincubated at 2 C (top) or competitive. The resulting Lineweaver-Burk (8) plots (Fig. 2) 37 C (bottom) for times indicated before assay. Each point repre- verified Pressey's (11) findings of noncompetitive inhibition. sents the mean of six replications. Standard error, 0.01 ,umole/hr. The value of 5.7 mm sucrose for the Km was in good agree- Note the difference in time units used for 2 and 37 C. ment with his values of 6.1 mm (11) and 5.3 mm (12). The next question examined was the dissociability of the enzyme-inhibitor complex. To investigate this, the concentration of enzyme was held constant, and the amount of added inhibitor was varied. The results were plotted in two ways: (a) with inhibitor concentration plotted against the reciprocal of the reaction rate, and (b) with inhibitor concentration plotted against percentage inhibition. If the complex is highly dissociable, the first method of plotting should produce a straight line and the second method should result in a pronounced curve. On the other hand, if the complex is undissociable, the first method should give a curve and the second a straight line. As shown in Figure 3, both methods of plotting yielded graphs 2 A consistent with an undissociable complex. Also consistent with 1/s x 10 this type of complex was the graph of percentage inhibition FIG. 2. Lineweaver-Burk plots for partially purified invertase plotted against enzyme concentration at a constant level of and inhibitor showing noncompetitive inhibition. Assays were per- inhibitor (Fig. 4). Percentage inhibition decreased with increas- formed after preincubating enzyme and inhibitor for 4 days at ing enzyme concentration, and the rate of decline was less the 2 C. Points are means of four replications. greater the amount of enzyme added. In contrast, an enzyme- 368 EWING AND McADOO Plant Physiol. Vol. 48, 1971 for example, by stirring vigorously in a blender or by bubbling N. through the extract. He concluded that the inhibitor was surface-denatured by the foaming and that the invertase was unaffected. In later work (13, 14) the blending technique, slightly modified, was utilized to estimate "total" and "basal" invertase levels in potato tuber extracts. However, Moll (9), using a different type of blending apparatus, was unable to destroy inhibitor without some loss of invertase activity. We, 1.0 | / 0 With inhibitor therefore, foamed extracts for various time periods by stirring in a blender and by bubbling N2. Vigorous stirring of tuber extract in a VirTis blender accord- 0.5 X ing to the method of Pressey and Shaw (14) resulted in increased invertase activity, but in our hands the method was 0 25 50 75 not selective. As shown in Figure 6, more than a few minutes Invertase, MI/sample of stirring appeared to produce inactivation of the enzyme. Foaming accomplished by bubbling N2 appeared to destroy the FIG. 5. Relation between aliquot size and reaction rate as af- inhibitor more slowly but was less destructive to the enzyme fected by presence or absence of inhibitor in the solution being (Fig. 6). When partially purified invertase was subjected to sampled. Concentration of partially purified invertase was the blending, enzyme destruction was even more apparent (Fig. 7). same in both solutions. Ratio of inhibitor to enzyme was constant In view of the unsatisfactory results obtained by stirring in in samples containing inhibitor, which were preincubated 2½/2 days at 2 C after dilution. the blender, we examined the effects of bubbling N2 through solutions of partially purified enzyme and inhibitor. Treatments and results from two separate experiments, each replicated six times, are presented in Table I. Foaming produced 3-fold in- 1.2- u creases in the invertase activity of enzyme-inhibitor mixtures, with no significant differences whether foaming was for 5, 10, or 20 min. Foaming also produced a slight increase in the 0.9 activity of partially purified invertase to which no inhibitor had been added. Apparently some inhibitor had survived the purifi- Blending cation procedure even though the first step in purifying was to 'E 0.6 stir for 30 min in the VirTis, compared to the 10-min blending recommended by Pressey (12). The only other significant difference in enzyme reaction rates occurred when invertase was 0.3 added after the foaming of the inhibitor. In this case con- taminating inhibitor present with the invertase would not have been foamed; and, as might be expected, values did not signifi-

0 20 40 60 cantly differ from that obtained for the unfoamed enzyme to Foaming time, minutes which no inhibitor was added. There was some indication that preincubation of enzyme and inhibitor protected the inhibitor FIG. 6. Comparison of two methods of foaming crude extract against inactivation, since after 5 of foaming the samples from potato tuber. Solid line shows effect on enzyme activity of min passing N2 through the extract for various periods. Broken line preincubated before foaming showed slightly higher reaction indicates results of stirring in VirTis blender. Volume taken for rates than did samples which were preincubated only after assay was 0.1 ml in each case. Standard error, 0.02 Amole/hr (three foaming. The difference disappeared with longer foaming replications). periods. Other experiments (data not shown) indicated no detectable inhibitor complex with a relatively high dissociation constant difference whether the water bath temperature was 27 or 37 C would produce the same percentage inhibition regardless of enzyme concentration (6). 2.5 Schwimmer et al. (16) were led to suspect the presence of No "blending" the inhibitor in crude tuber extracts because with increasing 1 amounts of extract assayed, reaction rate did not increase 2.0 min 2 min proportionately. Pressey (11) obtained similar results and con- sidered that a linear relation between amount of extract assayed 1.5 and enzyme reaction rate was evidence for freedom from inhibitor. However, this should not apply to an inhibitor forming an undissociable complex. In fact, as shown in Figure 5, there /4 is no more deviation from linearity when a mixture of enzyme 2 ^<. 8~~min and inhibitor is taken for assay than there is when partially purified invertase is assayed alone. If, as these results indicate, the enzyme-inhibitor complex is undissociable, or nearly so, then classical inhibition kinetics cannot be applied to determine the amount of enzyme and 0 0.2 0.4 0.6 inhibitor present in tuber extracts (4). Hence, it would be Invertase, ml/sample especially helpful to have some method of selectively inactivat- FIG. 7. Inactivation of partially purified invertase by stirring ing either the invertase or the inhibitor in a mixture of the two in a blender. Aliquots were removed at intervals after stirring for without affecting the activity of the other. Pressey (11) claimed indicated times and assayed for invertase activity. Points repre- that this could be accomplished by "foaming" potato extracts, sent means of duplicate determinations. Plant Physiol. Vol. 48, 1971 POTATO TUBER INVERTASE AND INHIBITOR 369 during foaming by either method. It was, however, important Table I. Reaction Rates of Invertase-Inhibitor Mixtures Subjected to control the gas pressure of the N2 to give a uniform rate of to Foaming before Mixinig, after Mixing but before Preincubation, bubbling. If the pressure was regulated to give approximately or after Both Mixinig and Preincubation four bubbles per sec, then the foaming time required to achieve Each sample assayed contained 0.1 ml of partially purified in- maximal enzyme activity was very reproducible for aliquots vertase (e). Partially purified inhibitor (i) was added in 0.2-mi of any given extract, although highly variable for extracts aliquots to treatments 2, 3, and 4. Volumes were adjusted to 0.5 from different potatoes. ml with buffer. Preincubation was 2 days at 2 C. Foaming was by bubbling N2. Reaction rates are means of six replications. DISCUSSION Reaction Rate with Foaming for: Preincubation The observation which led to the discovery of invertase Treatment inhibitor in potato tubers was deviation from linearity that 0 min 5 min 10 min 20 min occurred when reaction rate was plotted against crude enzyme concentration (11, 15, 16). Ironically, the inhibitor discovered jgnoles redzucing sugar/hr X 102 (12) is not of the type which would affect linearity. It is clear 1. e only 86 bcl 96 ef 97 ef 96 ef from the data presented in Figures 3, 4, and 5 that this in- 2. i foamed, e added, then pre- 84 b 87 bc 86 bc hibitor forms an essentially undissociable complex with potato incubated tuber invertase. Therefore, it would have no detectable effect 3. i + e, foamed immediately, ~29 a 95 ef 100 f 93 de on the linearity of reaction rate plotted against amount of en- then preincubated zyme. 4. i + e, preincubated, then 88 bcd 91 cde 93 de The reason for the deviation from linearity is not known. foamed, preincubated We, too, frequently observed a nonproportionality between again reaction rate and aliquot size of crude extract. To a lesser de- 1 Values followed by a common not gree, this was often true of invertase preparations partially letter do differ significantly purified by Pressey's (12) procedure. Whatever the explanation, at the 5%/3o level by Duncan's Multiple Range Test. it cannot be the presence of an inhibitor such as Pressey's (12) which forms an undissociable complex with the enzyme. but the experiments with partially purified enzyme showed that The low dissociability of the enzyme-inhibitor complex also it too was affected by the foaming. means that the methods developed by several groups of workers Foaming by passing N2 through the extracts at a controlled (5, 9, 16) for estimating invertase in the presence of inhibitor rate appeared more promising. The destruction of inhibitor was are inappropriate. All are based upon an equation by Ebersole less rapid than in the blender, but there was much less indi- et al. (3) which can be shown to be applicable only to reversible cation of invertase inactivation. In fact, the activity of partially inhibition of relatively high dissociation constant (2, 4). From purified invertase was increased by N2 foaming. This was con- the available data it is not possible to state whether the enzyme- firmed with other batches of enzyme, although the size of the inhibitor complex is completely undissociated (K, = 0) or response varied considerably from one batch to another. Pre- whether a very slight degree of dissociation occurs (very small sumably a small amount of inhibitor was present as a con- K,).' A plotting of the data from Figure 3 according to the taminant, even though one would not expect it to have survived equation for high affinity, reversible inhibition (2) produced 30 min of foaming in the blender only to be inactivated by considerable scattering from a straight line. There was no indi- 5 min of bubbling N2. cation that the slope of the line, which should be proportional The length of time required to remove all inhibitor by N2 to K,, was greater than zero. Whether the complex is com- foaming varied from extract to extract. Data in Table I indi- pletely undissociated or only nearly so, previous methods (5, cate that S min were sufficient to inactivate inhibitor in solu- 9, 16) based upon an equation which assumes a highly dissoci- tions prepared by mixing partially purified invertase and in- ated enzyme-inhibitor complex are not applicable (4). hibitor. The tuber extract represented by the data in Figure 6 Although Pressey (11, 12) did not mention the need for pre- seemed to require longer periods. Still other extracts reached incubation of enzyme and inhibitor in order to permit com- maximal invertase activity with foaming periods ranging from plex formation to reach equilibrium, we found, as he did, that 10 to 40 min. Such variation was not caused by random ex- Lineweaver-Burk plots indicate noncompetitive inhibition. perimental error, as the results for a given extract were much Thus the enzyme-inhibitor complex can be characterized as more reproducible than the results for different extracts. Thus noncompetitive with substrate, dependent upon temperature it would appear that approximations of relative inhibitor and with respect to rate of formation, and essentially undissociable. invertase concentrations can be obtained by assaying extracts Pressey and Shaw (14) showed that during storage of potato before and after the passage of N2 through the extracts. In tubers either invertase or inhibitor may be in excess and that following this procedure, it would be desirable to ascertain the high storage temperatures tend to shift the balance toward the optimal foaming period for each extract and to carefully con- inhibitor. It obviously would be of interest to follow tempera- trol bubbling rate. ture-induced changes in absolute concentration of both inhibi- Acknowledgments-We gratefully recognize the able assistance provided by tor and enzyme. At least in our hands, however, the method of Miss 'Margaret E. Lins and Miss Lourdes Dy in performing these experiments Pressey and Shaw (14) for selectively inactivating inhibitor by and helpful criticism given by Professors David C. Wharton and John F. Thomp- son during the foaming in a blender was not satisfactory. Such foaming did preparation of the manuscript. increase the invertase activity of crude extracts, indicating that LITERATURE CITED the inhibitor was inactivated more rapidly than the enzyme; 1. Dixox, M. 1953. The determination of enzyme inhibitor constants. Biochem. J. 55: 170-171. Since the molecular weight of the inhibitor has been only 2. DixoN-, M. AN-D E. C. WEBB. 1964. Enzymes. Academic Press, New York. pp. esti- 331-332. mated from Sephadex chromatography and the inhibitor has been 3. EBERSOLE, E. R., C. GuTTENTAG, AND P. W. WILSON. 1944. Nature of carbon only partially purified (12), a reliable value of K, cannot be com- monoxide inhibition of biological nitrogen fixation. Arch. Biochem. 3: 399- puted. 418. 370 EWING AND McADOO Plant Physiol. Vol. 48, 1971

4. EWING, E. E. 1971. Graphical determination of relative concentrations of en- 10. NELSON, N. 1944. A photometric adaptation of the Somogyi method for the zyme and endogenous inhibitor by dilution. Plant Physiol. 48: 371-372. determination of glucose. J. Biol. Chem. 153: 375-380. 5. FROST, G. M., R. N. GREENSIELDS, AND F. W. J. TEALE. 1988. Some prop- 11. PRESSEY, R. 1966. Separation and properties of potato invertase and in- erties of 83-fructofuranosidases partially purified from Phaseolus vulgaris vertase inhibitor. Arch. Biochem. Biophys. 113: 667-674. and Solanum tuberosum. Biochem. J. 107: 625-636. 12. PRESSEY, R. 1967. Invertase inhibitor from potatoes: Purification, character- 6. KERN, M. AND R. NATALE. 1958. A diphosphopyridine nucleotidase and its ization, and reactivity with plant invertases. Plant Physiol. 42: 1780-1786. protein inhibitor from Mycobacterium butyricum. J. Biol. Chem. 231: 41-51. 13. PRESSEY, R. 1969. Role of invertase in the accumulation of sugars in cold- 7. KUNiTZ, M. AND J. H. NORTHRUP. 1936. Isolation from beef pancreas of crys- stored potatoes. Amer. Potato J. 46: 291-297. talline trypsinogen, trypsin, a trypsin inhibitor, and an inhibitor-trypsin 14. PRESSEY, R. AND R. SEAW. 1966. Effect of temperature on invertase, invertase compound. J. Gen. Physiol. 19: 991-1007. inhibitor, and sugars in potato tubers. Plant Physiol. 41: 1657-1661. 8. LINEWEAVER, H. AND D. BuRK. 1934. The determination of enzyme dissocia- 15. ROREM, E. S. AND S. SCuWIMMER. 1963. Double pH optima of potato invertase. tion constants. J. Amer. Chem. Soc. 56: 658-666. Experientia 19: 150-151. 9. MOLL, A. 1969. Die Saccharase im Stoffwechsel der Kartoffelknolle. II. Unter- 16. SCHWIMMER, S., R. U. MAXOWER, AND E. S. ROREM. 1961. Invertase and in- schiedliche Hemmungskoeffizienten der Saccharase in Knollen verschiedener vertase inhibitor in potato. Plant Physiol. 36: 313-316. Sorten. Flora, Abt. A 160: 301-305. 17. SOMOGYI, M. 1952. Notes on sugar determination. J. Biol. Chem. 195: 19-23.