[Reprinted from THE JouRNAL of GENERAL PuysIoLocy, January 20, 1941, Vol. 24, No. 3, pp. 377-397] STUDIES ON THE LACTASE OF ESCHERICHIA COLI By H. P. KNOPFMACHER anp A. J. SALLE (From the Depariment of Bacteriology, University of California, Berkeley) (Received for publication, September 16, 1940) A yeast capable of fermenting lactose was first described by Adametz (1889). He found it in his studies on the microorganisms of cheeses and gave it the name Saccharomyces lactis. In the same year Beijerinck work- ing with two species of yeast, Saccharomyces kefir and S. tyrocola, succeeded in demonstrating in the filtrate of his cultures a lactose-hydrolyzing enzyme, which he named “‘lactase.” Following these investigations lactases were soon detected in many yeasts, molds, bacteria, and in animal tissues. In 1896 Fischer and Niebel voiced the opinion that hydrolysis had always to precede the fermentative decom- position of lactose. From their study of the structure of carbohydrates they concluded that the enzyme concerned must be specific for the alpha- glucose-beta-galactoside linkage of milk sugar. Due to more recent work, however, the validity of these assumptions has become rather questionable. Lactases are widely distributed in the plant and animal kingdoms. Euler (1922) in reviewing the literature on this subject points out that they are always found in the intestinal tract of young mammals but decrease mark- edly with age. As to their occurrence in the pancreas there is no agreement among the various authors. More recently Cajori (1935) has reported a lactase from the dog’s liver. Bierry and Ranc (1909) found a lactase in the gastrointestinal tract of the edible snail, Helix pomatia, and Wigglesworth (1927) reported it from the midgut of the cockroach, Periplaneta americana. It is, however, very doubtful whether these lactases are identical with those of higher animals, and the same holds for the lactases of higher plants, most frequently encoun- tered in the family Rosaceae. The best known example in this group is the enzyme emulsin of bitter almonds, which can hydrolyze lactose as well as beta-glucosides. Various species of yeas‘s, molds, and bacteria are capable of fermenting lactose and may contain lactases. Such have been found in Aspergillus niger and A. oryzae by Hofmann (19344), in Diplococcus pneumoniae by Fleming and Neill (1927 a), in Clostridium perfringens by the same authors 377 378 LACTASE OF ESCHERICHIA COLI (19270), in Escherichia coli by Lowenstein, Fleming, and Neill (1929), and in Escherichia coli mutabile by Hershey and Bronfenbrenner (1936) and Deere, Dulaney, and Michelson (1936). The presence of lactases in these organisms, however, does not necessarily mean that hydrolysis of the lactose into its constituent sugars has to precede fermentation. The evidence ob- tained by Willstatter and Oppenheimer (1922) for lactose yeast, by Wright (1936) for Streptococcus thermophilus, and more recently by Leibowitz and Hestrin (1939) for maltose yeast points very strongly to the possibility of direct fermentation of lactose and other disaccharides under certain conditions. Escherichia coli was selected for a general study of its lactase, special emphasis being placed on the kinetics of enzyme action, heat inactivation, and the behavior of the enzyme toward some reducing and oxidizing agents and salts of heavy metals. EXPERIMENTAL f. Preparation of the Enzyme Solution Fleming and Neill (1927) were successful in obtaining cell-free extracts of carbo- hydrases from pneumococci by subjecting them to repeated freezing and thawing. In this process zymases were destroyed, while the activity of the hydrolytic enzymes was preserved. This method is very tedious and time-consuming and ‘therefore was “not investigated further. Hofmann (1934 6, c) obtained active lactase preparations from EF. coli and B. del- brukit by treating the bacteria with an alcohol-ether mixture and then drying them at room temperature. This method was found to be unsatisfactory in our hands largely because of the susceptibility of the enzyme itself to the solvents used. The activity of preparations of lactase so obtained was very low and decreased on prolonged contact with alcohol, which sometimes was unavoidable. To obtain appreciable amounts of enzyme, masses of E. coli were grown on standard meat extract agar in which 1.5 per cent of lactose had been incorporated. After 48 hours incubation the organisms were washed off with a physiological salt solution, con- taining 1 per cent of toluene, and subsequently centrifugated. This procedure was repeated three times in order to reduce to a minimum the concentration of adhering metabolic waste products. The resulting suspension contained 6 X 10! organisms per ml. It was treated with an additional amount of toluene bringing the total con- centration of the latter up to 5 per cent. Toluene serves three purposes: (1) It acts as a preservative, (2) it inactivates the zymase complex without affecting the lactase (Willstatter and Oppenheimer, 1922), and (3) it destroys the semipermeability of the cell walls, bringing about a gradual autolysis of the bacteria. — Several attempts were then made to obtain a cell-free enzyme preparation. As was mentioned above it was found that the lactase apparently was very susceptible to alcohol and ether. It was also completely inactivated on dehydration with acetone. When a toluene-treated cell suspension was incubated overnight at 37°C. a dry gelatinous sub- stance was obtained. This was removed and ground to a powder, the relative activity H. P, KNOPFMACHER AND A. J. SALLE 379 of which, as determined by a method to be described later, was found to be 82 per cent of that originally present in the bacterial suspension. The dried powder, consisting of whole cells and cell fragments, was subjected to more rigorous autolysis. Measured portions were suspended in m/15 phosphate buffer solutions of pH 7.0, 8.0, and 9.0 and incubated overnight at temperatures of 37°C. and 46°C. They were then centrifugated and supernatants and sediments tested for lactase activity. The opaque supernatant fluids were practically inactive, whereas the precipitates still exhibited a marked activity though less than that of the dry powder, probably because of the severity of the treatment to which they had been subjected. A microscopical examination revealed that practically all bacterial cells were disinte- grated, and only cell fragments were present. The enzyme, apparently, adhered to these cell fragments. These observations are contrary to reports by Karstrém (1930), who obtained cell- free lactase preparations from &. coli by suspending the dried organisms in phosphate buffer solution of pH 7.0. They are, however, in agreement with results reported by Hershey and Bronfenbrenner (1936), who were unable to separate the enzyme from the bacterial cell and therefore concluded that it was an intracellular water insoluble enzyme. In another experiment equivalent amounts of toluene-treated cell suspension were exposed to the action of trypsin and papain. In both instances lactase activity was destroyed. Finally, 120 ml. of bacterial suspension were ground for 18 hours in a ball mill devised by Krueger (1933). But again lactase was inactivated. In view of these experiences it was decided to use the original cell suspension in all subsequent experiments, and it will be referred to in this report as “enzyme solution” or “E. coli lactase” inasmuch as it was solely employed for hydrolyzing lactose. This preparation was stored in an icebox at 5°C. where its activity decreased only slightly during the course of several months. 2. Materials and Methods Standard sugar solutions: 1 gm. of lactose hydrate and glucose, respectively, were dissolved in 100 ml. of distilled water and a few drops of toluene added. Throughout the course of the experiments dilutions were prepared from these standard solutions, 1 ml. of which contained 10 mg. of the respective sugar. The Folin-Wu method (1920) for blood sugar determination was chosen as best fitted for measuring the total amount of sugar present before and after hydrolysis by the enzyme. Experiments were conducted as follows: The desired dilution of the standard was prepared by the use of M/15 phosphate buffers of measured hydrogen jon concentration. One-tenth ml. portions of enzyme preparation were added to 5 ml. of lactose solution and the tubes shaken in a water bath at 36°C. for a certain length of time. Thereupon, they were centrifugated for 30 minutes and the supernatant liquid used for sugar deter- mination. 2 mi. were pipetted into Folin-Wu sugar tubes, 2 ml. of copper solution added, and the tubes then placed in boiling water for 8 minutes. After cooling 2 ml. of color reagent (phosphomolybdic acid) were added, the tubes made up to a volume of 25 ml. with water, and the resulting color compared with that of a standard. In preliminary readings, employing glucose and lactose solutions of different concen- trations, it was found that 1 mg. of lactose corresponded to 0.504 mg. of glucose. In 380 LACTASE OF ESCHERICHIA COLI all experiments, therefore, the values obtained have been expressed in terms of glucose or total reducing sugar on the basis of the above empirical determination. For example, if the initial concentration of lactose is 1/40 of that of the standard solution, i.c. 2 ml. contain 0.5 mg. of lactose, it will be read as 0.252 mg. of glucose or total reducing sugar, a glucose solution being always used as the standard for com- parison. For each experiment a parallel control had to be set up since most. of the chemicals whose effect on the enzyme was to be tested were oxidizing or reducing agents, and the enzyme solution itself slightly reduced copper sulfate. For this purpose corresponding amounts of enzyme and chemical reagent were added to 5 ml. of phosphate buffer solu- tion and the reducing values obtained then subtracted from the total. Finally, a correction for volume had to be made to an extent dependent upon the amount of enzyme solution and chemical reagent added. It was impossible to maintain a perfectly uniform rate of hydrolysis for the duration of the experiments. The values fluctuate between 51 and 59 per cent hydrolysis per hour for a 1/40 lactose solution. This circumstance, however, was not regarded as of importance inasmuch as the problem selected concerned merely the comparative study of rates of reaction as affected by hydrogen ion concentration, temperature, and chemicals. RESULTS 1. The Effect of Hydrogen Ion Concentration Optimal conditions with regard to hydrogen ion concentration differ for lactases from various sources (Oppenheimer, 1935). To determine the effect of pH on the activity of EZ. coli lactase, experi- ments were carried out as follows: m/i5 phosphate buffer solutions of different pH were prepared and their hydrogen ion concentration checked by means of a glass‘electrode. They were then used to make up lactose solution of a concentration of 1/40 with respect to the standard (0.252 mg. of total sugar per 2 ml.). As described above, 5 ml. were then mixed with 0.1 ml. of enzyme prep- aration and shaken in a water bath at 36°C. for 1 hour, and the reducing sugar was determined. The results are given in Table I. The values are plotted in Fig. 1. The results indicate that the activity of the enzyme is markedly reduced by slight acidity but much less affected by alkalinity of the medium. The optimum pH for the time period and temperature given seems to extend over the range between 7.0 and 7.5. Consequently, all subsequent experi- ments were carried out at a pH of 7.5. 2. The Mechanism of Enzyme Action Michaelis and Menten (1913) worked out general rate laws for the action of invertase on sucrose by assuming a chemical combination of the enzyme with its substrate as the governing step in the hydrolysis of the sugar. H. P. KNOPFMACHER AND A. J. SALLE 381 The enzyme-substrate equilibrium can be represented by the equation: _ BS) (és) TABLE I Effect of pH on the Degree of Hydrolysis of Lactose by E. coli Lactase K, Ratio of activity to that pH Amount of total sugar Hydrolysis of maximum activity mg. per 2 ml. per cent 5.0 0.254 0.8 0.01 6.0 0.357 41.7 0.71 6.5 0.386 53.2 0.91 7.0 0.399 58.3 0.99 7.5 0.400 58.7 1.00 8.0 0.393 56.0 0.95 9.0 0.381 §1.2 0.87 OO 90] 8 ao wy = 960} R \ | Percent h & 8 PS N\ 2 [ ; [ 5 6 pH 7 8 9 Fic. 1. Effect of pH on the rate of hydrolysis of lactose by £. coli lactase. where (£) and (ES) refer to the concentration of free and combined enzyme respectively and (S) to the concentration of the substrate. The constant k, could be determined by simple mathematical calculation, leading to the equation v (S) Van Va b+ orem ©) (% -1) 382 LACTASE OF ESCHERICHIA COLI in which v represents the initial velocity at the substrate concentration (S), Vm the maximum velocity, %,, therefore, being equivalent to the sub- strate concentration at which half the limiting velocity is reached. Lineweaver and Burk (1934) developed graphic methods for determining dissociation constants of enzyme-substrate compounds. Since in some cases one molecule of enzyme reacts with several molecules of substrate they modified the Michaelis-Menten equation accordingly: », - Bsr * (ESn) and ov (sr Vm (S++ he The latter equation can then be written 1 k, 1 2 Vals TV, in which V,,, the maximum velocity, and &, are constant. A plot of i against = must therefore give a straight line for some in- v n : . os i .. 1 : tegral value of ~. The intercept of this line on the - axis is —— and its v m slope a In this fashion, then, the constants are easily determined. When the above equation is multiplied by (S)" it assumes the form (Stk, 0 Va Va obtained. The intercept on the By plotting (S)" against (S)* a straight line is again v (S)* 2 axis is x, and the slope is > The latter plot is not only of importance in checking the values obtained by the former but also in discovering any departure from a straight line due to substrate inhibition. In such a case plots of sy against (S)" give curves that rise concavely with increasing substrate concentration. The following solutions were prepared: (a) A 1/10 dilution of the standard (2 mg. of lactose per 2 ml. = 29.3 X 10-*m) (b) A 1/20 dilution of the standard (1 mg. of lactose per 2 ml. = 14.6 x 10-4m) H. P. KNOPFMACHER AND A. J. SALLE 383 (c) A 1/40 dilution of the standard (0.5 mg. of lactose per 2 ml. = 7.3 X 10-4ur) (d) A 1/60 dilution of the standard (0.33 mg. of lactose per 2 ml. = 4.9 xX 10-‘m) The results of hydrolysis after 30 and 60 minutes are given in Table II. Upon plotting 1/» against 1/S and S/v against S practically straight lines were obtained. (See Figs. 2 and 3.) Consequently, it can be concluded that one molecule of enzyme combines with one molecule of lactose as is the case with all the other carbohydrases so far investigated. TABLE II Rate of Hydrolysis of Varying Concentrations of the Substrate Amount of total | Velocity 1/9 (av.) 1/S S/o mg. per 2 mi, per min, 1. Substrate A 30 min. 1.164 0.0053 192 0.5 384 60 “ 1.312 0.0051 2. Substrate B 30 min. 0.628 0.0041 260 1.0 260 60 “ 0.722 0.0036 3. Substrate C 30 min. 0.332 0.0027 392 2.0 60 “ 0.394 0.0024 196 4, Substrate D 30 min. 0.233 0.0022 500 3.0 167 60 “ 0.273 0.0018 : The intercept on the 1/2 axis is at 138, hence V, = 1/138 = 0.0072 mg. per 2 ml. per minute. K, = Vm X slope = 0.0072 K 132 = 0.95 mg. per 2 ml. = 13.9 X 1074w or 0.00139 V,, and k,, as evaluated from the second plot, are somewhat higher. The intercept on the S/v axis is at 132 = k/Vm. 1/Vm = 130, hence Vm = 0.0077 and k, = 1.02 = 14.9 x 10-‘m, or 0.00149. From these plots it may be inferred that the substrate has no inhibiting effect on the rate of hydrolysis by the enzyme. To determine the effect of different concentrations of E. coli lactase on the rate of hydrolysis of lactose an experiment was set up in the ordinary way, using a 1/40 lactose solution but adding varying amounts of enzyme preparation. The results are recorded in Table ITI. Plotting the figures of the third column against those of the first gives practically a straight line (Fig. 4). This may be also expressed in a mathe- 384 LACTASE OF ESCHERICHIA COLI Va > i Ft oO is Ys 20 25 30 Fic. 2. Nature of the enzyme-substrate intermediate of Z. coli lactase with lactose. 500 i a 5 10 15 20 Fic. 3. Test of enzyme-lactose intermediate. H. P. KNOPFMACHER AND A. J. SALLE 385 matical form by the equation: K = 7 where x represents the amount hydrolyzed, E the enzyme concentration, and ¢ the time, which in the above experiment was constant; viz., 1 hour. It is at once evident that TABLE II Hydrolysis by Varying Enzyme Concentrations Schiitz constant Amount of enzyme | Amount of total sugar |Amount hydrolyzed (x) K= = Ki= Va oel, mg. per 2 ml. 0.05 0.319 0.067 1.34 0.30 0.10 0.383 0.131 1.31 0.41 0.15 0.452 0.200 1.33 0.52 0.20 0.499 0.247 1.24 0.53 KO 90 4 80 ia 3 70 ZL 8 zo £ 2 50 5 2 @ 40 2, 20 LZ 10 a05 “alo Ols . azo Enzyme concentration (in mi) Fic. 4. Relation between enzyme concentration and the rate of hydrolysis. the values for K fit the above data far better than those for Ky, the so called Schiitz constant (1885). Analogous results with yeast lactase were reported by Willstaétter and Oppenheimer (1922). In view of the fact that, as has been shown previously, a definite equilibrium between enzyme and sub- strate is established the products of reaction apparently do not decrease the rate of hydrolysis. 386 LACTASE OF ESCHERICHIA COLI 3. Kinetics of Lactose Hydrolysis by E. colt Lactase The hydrolysis of lactose by E. coli lactase follows a course between a zero and first order reaction which is quite common for hydrolytic enzymes. Michaelis and Menten (1913), confronted with such difficulty in the a3 3 I 7 / go Z| =< % a & Substrate ~ concentration yw I 293x104 M S I i46x10*M 3 I B73 x 104M = W49 x 10+M ¥ al 7 X i Oe 30 60 90 t2o Time (minutes) Fic. 5. Relation between substrate concentration and the rate of hydrolysis. case of invertase, showed that its action could be expressed by a formula that is actually a combination of zero and first order equations. Zero order: k = kof a First order: In ——— = &yf, a-% a Michaelis-Menten equation: Vat = «+ In —— Essentially the same formula was derived by Van Slyke and Cullen (1914) for the action of urease on urea. Barendrecht (1913) showed that it also held for lactase prepared from yeast. As suggested in the original paper of Michaelis and Menten, the data of Table II have been presented graphically in two ways. In Fig. 5 the amount hydrolyzed (x) is plotted against time (é). It is readily seen that H. P. KNOPFMACHER AND A. J, SALLE 387 for the highest concentration a straight line is obtained indicating that the zero order reaction holds in this case which is in agreement with Michaelis and Menten’s observations. a In Fig. 6, x + 2.3 &, log a-x is plotted against time, and practically straight lines result, at least for the Ist hour. Only in the case of the 1/40 lactose solution were additional data for the 2nd hour available (x = 0.185 20, 13 4 » Substrate al; I corcentration —# p10 I I 49x10%M 2 HZ 73xi0*M % I (4exio*M x WY 293x10'M bal Hl \ + PF M o. oo 30 60 90 120 Time (minutes) Fic. 6. Kinetics of the hydrolysis of lactose by £. coli lactase. after 1.5 hours and x = 0.214 after 2 hours), and these, too, follow prac- tically a straight line. Hence, one can draw the conclusion that hydrolysis of lactose by E. coli lactase approximates the reaction course of the in- tegrated Michaelis-Menten equation. 4. The Effect of Temperature 5 ml. of a 1/40 lactose solution were incubated with 0.1 ml. of enzyme preparation for 30 minutes at 26°C., 36°C., 46°C., and 56°C. after preheat- ing the enzyme for 5 minutes at the respective temperature. Table IV shows the results obtained. 388 LACTASE OF ESCHERICHIA COLI The recorded drop of lactase activity between 36°C. and 56°C. may be attributed most probably to a more rapid heat inactivation of the enzyme at the higher temperatures. To elucidate this point further a few experiments were set up designed to determine the rate of enzyme destruction at different temperatures. Test tubes containing measured amounts of enzyme solution were immersed in a water bath at the desired temperature which was closely controlled. TABLE IV Effect of Temperature on the Degree of Hydrolysis of Lactose by E. coli Lactase Temperature Amount of total sugar Hydrolysis | Ratio of activity °C. mg. per 2 mi. per cent 26 0.295 17.1 36 0.330 oe Ue 46 0.379 50.4 0 56 0.249 0 TABLE V Rates of Heat Inactivation of E. coli Lactase at Different Temperatures Température of Time of preheating | Amount of total sugar | Amount hydrolyzed | First order constant °C. mg. per 2 ml, 45 0 0.379 0.127 15 0.354 0.102 0.0146 20 0.345 0.093 0.0156 30 0,332 0.080 0.0154 53 0 0.379 0.127 3 0,336 0.084 0.138 5 0.319 0.067 0.128 7 0.300 0.048 0.139 After heating for varying times the tubes were placed in ice water to check as quickly as possible further destruction of enzyme. The residual lactase activity was then determined in the ordinary way of mixing 0.1 ml. of en- zyme preparation with 5 ml. of a 1/40 lactose solution and shaking it in a water bath of 36°C. for 1 hour. The results of experiments carried out at temperatures of 45°C. and 53°C. are given in Table V. It was found that thermal inactivation of E. coli lactase followed the equation of a simple first order reaction: 2.3 log Ao/A = #, H. P. KNOPFMACHER AND A. J. SALLE 389 where A, is the activity of the unheated enzyme solution, in other words, the amount hydrolyzed under ordinary conditions, A the activity of the enzyme heated for the time #, and & the constant of heat inactivation. The average values for & are thus 0.0152 at 45°C., and 0.135 at 53°C. They can be determined also by plotting log A against ¢ as has been done in Fig. 7. It is at once evident that the rate constant for heat inactivation changes -OF, +10)