ENZYMATIC HYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES AS MEASURED BY CHANGES IN VISCOSITY BY HILLEL S. LEVINSON AND ELWYN T.

REESE

(From the Quartermaster General Laboratories, Philadelphia)

(Received for publication, January 23, 1950) In a previous report (15), a method for the rapid assay of the hydrolytic activity of cell-free filtrates was reported, employing carboxymethyl cellulose Na salt (CMC) as the substrate. The assumption was made that the enzyme, Cx, hydrolyzing the 1,4-/3-glucosidic linkage in CMC is identical with that attacking the same linkage in cellulose, and that this enzyme is probably only one of those found in the cellulase complex. It was noted that a decrease in viscosity accompanied the production of reducing sugars. The present study was undertaken in an attempt to find out the relationship between viscosity changes and the formation of reducing sugars by filtrates. The viscosimetric method has previously been employed in the study of such enzymes as pepsin (1), amylase (2), hyaluronidase (6), and polymetaphosphatase (7, 8). Hydrolysis of soluble celluloge derivatives has also been studied in a preliminary manner by measuring changes in fluidity. Ziese (18), in a comparison of the cellulases obtained from snail gut and from malt, inactivated the animal cellulase but could not inactivate malt cellulase by the same treatments. From this, he concluded that the two enzymes were not identical. Letzig (10, 11) used the viscosimetric technique in the detection of thickening agents found in various commercial preparations, amongst which were some of the soluble cellulose derivatives. Reducing sugars were detectable on enzymatic hydrolysis of carboxymethyl cellulose (CMC), but not on hydrolysis of methyl cellulose. In the present paper, an attempt is made to evaluate the mode of CMC hydrolysis by cell-free filtrates on the basis of viscosity measurements and reducing sugar analysis, since qualitatively similar results in the rate of hydrolysis as determined by both methods at different pH values, temperatures, and degrees of substitution of the CMC, might constitute preliminary evidence for the action of only one enzyme.

Methods The cell-free enzyme solutions were prepared by filtering cultures of each organism through sintered glass. Merthiolate (to give 0.01 per cent) was added to all solutions as a preservative. The concentration of filtrates varied with the test being made. In general, one part of filtrate plus nine of substrate was used when measuring reducing 601

602

ENZYMATIC HYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES

dtsp d~

FILTRATE CONCENTRATION,AS % OF REACTION IIXTURE FIG. 1. Effect of filtrate concentrate

Filtrate: K 2 = Actinomyces sp. QM B814. The filtrate was obtained by growing the organism on hydroxy ethyl cellulose, for 10 days. H 26 = Aspergillus fumigatus QM 45h. The filtrate was obtained by growing the organism on cotton duck for 26 days. AM-X = Myrothecium verrucaria QM 460. Filtrate from growth on cotton duck in shake flasks.

H. S. LEVINSON AND l~. T. REESE

603

.9 K~

.6

E.7 E w u).6 O O -I 0. 5

z_
O lal et

.i

O ~ fT

2 4 6 8 I0 FILTRATE GONGENTRATION. AS % OF REAGTION MIXTURE

[on on rate of hydrolysis of CMC.

Conditions: Fluidity test: CMC 70M 1 per cent; citrate pH 5.0, ~/20 in reaction mixture.
Temperature 35C.

Reducing sugar test: CMC 50T 0.5 per cent; citrate pH 5.0, M/20 in reaction
mixture. Temperature 50C. Time 1 hour.

604

ENZYMATIC HYDROLYSIS OF SOLUBLE

CELLULOSE

DERIVATIVES

sugars, and one part per hundred when measuring viscosity changes. The reducing sugar tests were conducted at 50C.,--the viscosity tests at 35C. The low concentration and low temperature used in the viscosity measurements were selected in order to reduce the rate of fluidity change to a convenient level for observation. Citric a c i d - NaOH buffer (~/20) was used throughout. For part of the work it was thought desirable to use two different CMC preparations. In the reducing sugar test, CMC 50T was used because larger amounts of reducing sugars are produced from it than from CMC preparations of higher degreeof substitution. In the viscosimetric method, CMC 70M was used because it combined two desirable qualities. I t formed a clear uniform solution of suitable viscosity and an appreciable and measurable amount of reducing sugar was formed. In the later experiments, a 1 per cent solution of CMC 70M was used for both measurements. Reducing sugars were determined as glucose by the dinitrosalicylic acid (DNS) method of Sumner (17). The viscosity was determined by a Hoeppler viscosimeter at a temperature of 35 ~: 0.5C. The fall-times of the ball ranged from 30 to 250 seconds. The viscosity, 7, in centipoises was corrected for the specific gravity of the solution in the usual manner. Specific viscosity, ~,~, or that part of the viscosity due to the substances in solution, is defined,
~Tsolution ~ ~Tsolvent ~solvent r]~olution ~Tsolvent

while specific fluidity, ~,p is the reciprocal of v,p- The intrinsic viscosity [,1] is defined
as

[7] ~- ~'p as c approaches zero

C

(where c = concentration of solute in grams per 100 ml. of solution). The course of the hydrolysis has been plotted as the specific f l u i d i t ) (.1 x vs. time y of incubation (in minutes) since during the first few hours of hydrolysis, this relation1 ship is a straight line function. In our results, the value of - ~
(i.e. rate of change

of fluidity) is used as a measure of the enzymatic activity. Such values are comparable only when the initial specific viscosities are the same. Fortunately, this has been true in most of the experiments performed. When different batches or different concentrations of CMC are used, a correction must be employed. Ingelman and Malmgren (7), studying polymetaphosphatase, believe that a correction can be made by multiplying the rate by the specific viscosity at zero time. Hultin (6), studying hyaluronidase, employs a factor which takes into account only the substratum concentration (Table I). At any rate, a clearer picture is necessary of the correction factor to be applied to the rate of change of fluidity. In the work that follows, the procedure is standardized in such a way that the initial spc.cific viscosity remains the same. Where this has been impossible, the data have been subjected to interpretation based on rate, with and without the correction factors discussed above. The intrinsic viscosity is proportional to the average degree of polymerization

H. S. LEVINSON AND E. T. REESE

605

(DP) of a substance within certain limits of molecular size. The application of viscosity data to a determination of molecular weight is complicated by many factors, and is best left to those familiar with the problem. But the proportionality between DP and [~] is useful in estimating relative changes in DP during the course of hydrolysis. Intrinsic viscosity was determined in 4 per cent NaOH solution. Samples were removed from the incubating enzymc-substrate mixture at various times and the action halted by the addition of alkali.
RESULTS

I. Effect of Various Factors on Enzyme Activity A. Effect of Enzyme Concentration.--The effect of enzyme concentration on
rate of hydrolysis of CMC was followed for filtrates of three organisms. From the effect of enzyme concentration on the production of reducing sugars (Fig. 1), it is noted that with the Myrothecium verrucaria and Actinomyces sp. filtrates, the increase in production of reducing sugars is a direct function of the enzyme concentration for the lower portion of the range of concentrations. The Aspergillus fumigatus filtrate has a distinctly different type of curve, characterized by a rather sharp initial rise in reducing sugars, followed by a levelling off at relatively low filtrate concentrations and resulting in a crossing of the other two curves. Subsequent experiments have shown that this reversal in relative activity of the filtrates is an artifact. When the time of incubation (assay) was reduced from 1 hour to 30 minutes, the Aspergillus fumigatus filtrate retained its position as the most acti,Ce for a// concentrations of filtrates. The explanation lies in the observation that filtrates of this organism acting on CMC give curves which flatten out as the reducing sugar values approach 0.4, a value which is far below the limiting values for other filtrates. A true comparison of activity can be made only at reducing sugar levels appreciably lower than that value. Although this early levelling off is not as marked in the fluidity change curve (Fig. 1), as in the reducing sugar curves, it is apparent, Myrotkecium verrucaria filtrates exhibit a constant rise in the rate of change in fluidity as the enzyme concentration increases, over the whole range of enzyme concentrations used; Actinomyces sp. exhibits this constancy over the lower 60 per cent of the range; while AspergiUus fumigatuz filtrate shows its characteristic sharp rise in the lower 20 per cent of the range and a tendency to level off in the rest of the range. In general, the rate is a linear function of the filtrate concentration for values between 0 and 140 X 10-~. At higher filtrate concentrations, the curve deviates from the straight line. B. Effect of pH.--In the determination of the effect of pH on enzyme activity, filtrates of Aspergillus fumlgatus and Actinomyces sp. were tested in a series of CMC mixtures containing citric acid--NaOH buffer. The pH change during the course of the hydrolysis was negligible, as was also the change in initial viscosity, due to different hydrogen ion concentrations (3). A comparison (Fig. 2) reveals a striking similarity in the two sets of results.

606

ENZYMATIC HYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES

I 10-4 I ! 1 I

IS

14

12-

I(

a-

I 3

I 4

1 5

I 6

r
7

H26

pH

F ~ . 2. Effect

Filtrates:
Filtrate No. Organism Sub*trate Incubation Time days

K3 AA X AH 13 H 26

Aainomyces sp. QM B814 Actinomyces sp. QM B814 AspergiUusfumigatus QM 45h Aspergillusfumigatus QM 45h

Cotton duck Cotton duck Cotton duck Cotton duck

10 Combined, 9-18 13 26

II.

S. LEVINSON AND E. T. REESE

607

pH
on activity of Cx. Conditions: Fluidity test: CMC 70M 1 per cent; citric acid-NaOH buffer /20; filtrates 1:250 in reaction mixture. Temperature 350 C. Reducing sugar test: CMC 50T 0.5 per cent; K-PO4 buffer 1/10; filtrates 1:10 in reaction mixture. Temperature 50C.

608

ENZYMATIC HYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES

According to the reducing sugar data, the pH optimum for the Actinomyces sp. filtrate occurs between 6.4-6.8. The pH optimum as determined in the fluidity test does not have quite so broad a range and occurs at 6.8-7.2. Likewise, by both types of assays, the Actinomyces sp. filtrate is almost inactive at pH 4.0, while it still remains comparatively active at pH 8.0. For the Aspergillusfumigatus filtrate, reducing sugar data indicate that the pH optimum is 4.7-5.1. At 4.0, the activity is considerably greater than that at pH 7.0 and approaches complete inactivity (by extrapolation) at pH 8.0. Fluidity test data similarly indicate a pH optimum at 4.6-5.0 with high activity at pH 4.0, and almost none at pH 8.0. The two filtrates reported here were selected to represent organisms having widely different pH optima for this hydrolysis. Several other fungi which have been studied show optima between 4.5 and 6.0. In some cases, the curves show a sharp peak, in others, a rather broad zone of high activity. The pH optimum for enzymatic hydrolysis thus appears to be related to the organism from which the filtrate is derived. C. Effect of Temperature.--In general, the activity of the filtrates, measured by both fluidity and reducing sugar methods, increases as the temperature increases (Fig. 3). In both series of experiments, two of the filtrates, those of Actinomyces sp. and Aspergilkus fumigatus, showed maximum activity above 60C. while the third, that of Myrothecium verrucaria, showed maximum activity at 53C. The falling off in activity of the Myrothedum verrucaria filtrates in the presence of substratum may be due to the inactivation of the filtrate at the higher temperature. In any case, the curves for fluidity and for reducing sugars are similar to each other with respect to the rather sharp temperature optimum for Myrothecium verrucaria and with respect to the absence of an optimum below 62C. for the other organisms tested. While the rate of change of fluidity increases as the temperature increases, a constant Q10 is' not observed over the range covered. The highest value observed is that for Myrothecium verrucaria (3.7), a value which is about double that found for the other two organisms (1.7, 1.8) for the same range (35-45C.). The Q10 falls as the temperature increases. This appears to indicate temperature inactivation either directly on the enzymes involved, or on other principles not determined (e.g. chemical inhibitors or activators). Q10 has little meaning in the reducing sugar measurements since an accurate comparison of rates has not been made. D. Effect of Substratum Concentratlon.--As was pointed out in the introductory paragraphs, the values of dl */'pare comparable within an experiment only

dt

when the initial specific viscosities are the same. In all the experiments thus far reported, this has been the case. However, occasions do arise in which the

H. S. LEVINSON AND E. T. REESE

609

initial ~,p values differ considerably. Such is the case when different batches, or when different concentrations of CMC are used (Fig. 4). The uncorrected rate of change of specific fluidity increases approximately fivefold as the concentration of the substratum decreases fourfold. However, if either Ingelman's or Hultin's factor is applied (Table I), a narrowing of the range of variation to 2 to 3 times results, together with a direct correlation between rate of change of fluidity and substrate concentration. The present data are insufficient to permit us to determine whether the corrections or the experimental techniques are inadequate. E. Effect of Degree of Substitution.--In previous work (15) it was shown that the degree of substitution (DS) had a decided effect on the ease of enzymatic hydrolysis of CMC as measured by the production of reducing compounds. It is apparent that the same trend is shown by the fluidity data (Fig. 5). In both cases the CMC is more readily attacked when it is not highly substituted. The data indicate that a degree of substitution involving one substituent per anhydroglucose unit is sufficient to render the CMC refractory to enzymatic hydrolysis. At DS 1.2, a slight change in the viscosity does take place whereas no reducing groups are measurable. It is believed that this is not significant, being merely a limitation of the end-group measurement used.

H. Effect of Various Factors on Enzyme Stability A. Effect of p//.--Several filtrates were subjected to inactivation by incubation at the desired hydrogen ion concentrations for 24 hours at 50C. The mixtures contained 9 mh of filtrate and 1 ml. of ~r/2 citrate buffer. The similarity in the two sets of data is striking, leading once more to the appearance of the identity of the enzymes involved in reducing sugar production with those involved in fluidity changes. For instance (Fig. 6), the pH of maximum retention of activity of the Myrothecium verrucaria filtrate was about 5.2 by the reducing sugar method and the same by the fluidity method. The curves for the Aspergillus fumigatus and for the Actinomyces sp. filtrates show a broad pH range in which there is little loss of activity as measured by both the sugar and viscosity methods. The Myrothecium filtrate has a rather narrow range in which the maximum activity is retained, the activity falling off rapidly at hydrogen ion concentrations above and below the pH of minimum inactivation. The pH inactivation curves indicate that the results pointed out in the section on pH relating to activity of the enzyme (Fig. 2) cannot be accounted for by inactivation of the enzyme, but are rather actual effects of the pH on the activity of the filtrates. Thus, the activity of the Actinomyces filtrate is almost nil at pH 4.0 with a short period of incubation, while the same filtrate retains approximately 80 per cent of its activity when kept at that pH level for 24 hours. Similarly, the Aspergillus filtrate has practically no activity

610

ENZYMATIC H Y D R O L Y S I S OF S O L U B L E C E L L U L O S E D E R I V A T I V E S

II

II

88 .x10-4

45

85 dt 25 AM X

AX

15

0 ',0

I 50

I I 40 50 TEMPERATURE C.

I 6O FIG. 3. Effect of tempei

Filtrates:
Filtrate designation Organism Days of growth on cotton duck

H 26 AH 13 AM X AM 13 AA X AA 13

Aspergillusfumigatus QM 45h Aspergillusfumigatus QM 45h Myrothecium vo'rucaria QM 460 Myrothedum verrucaria QM 460 Actlnomyces sp. QM B814 Aclinomyces sp. QM B814

26 13 9, 18 combined 13 9, 18 combined 13

H. S. L E V I N S O N

AND

E. T. R E E S E

611
I

.7 ..6

AI3

i,

!.5

.2 I 60

TEMPERATURE (='C~
a rate of hydrolysis of CMC.

Condition~: Fluidity test: Buffer: citric acid-NaOH, pH 5.0, M/20 in final mixture. CMC 70M 1 per cent; filtrate concentration 1:100 for AM X and AA X; 1:250 for H 26. Time 50 to 70 minutes. Reducing sugar test: Buffer: KH2PO4-K~HPO4, pH 4.5, M/20 in mixture. CMC 50T 0.5 per cent; filtrate concentration 1:10; time, 1 hour.

612

ENZYMATICHYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES

| I I

64

__X i0-2

0.25 % CMG

56

48

40

Jll P32
2d

0 . 5 0 % CMC

J
TIME OF INCUBATION (Hours)

I.O0 % CMC

Fro. 4. Effect of substrate concentration on slope of fluidity curve during enzymatic hydrolysis.

Filtrate: Aspergillusfumigatus QM 45h, grown on cotton duck 26 days. 1:1000 in

reaction mixture.

Conditions: Buffer: Citric acid-NaOH at pH 5.0, ~r/20. Substratum: CMC 70M.

against CMC at p H 8.0, while the same filtrate loses only about 10 per cent of its activity when maintained for 24 hours at that p H level. But the conditions in the two tests (activity, inactivation) are not identical. I n the activity tests,

H . S. L E V I N S O N AND E. T. R E E S E

613

the substratum is present during the test, while in the inactivation experiment the substratum is absent. The rate of inactivation of the enzyme in the presence of substratum may therefore be somewhat different from the results' being reported. B. Effect of Temperature.--Further comparison between results obtained by the two assay methods was made (Table II). Filtrates having Cx enzyme activity were adjusted to pH 5.2 with citrate buffer and incubated in a water bath at 70C. Samples removed at 30 and 120 minutes were immediately cooled and assayed for activity. In the reducing sugar method, an enzyme dilution curve was made from the original filtrate and the residual potency of
TABLE I

Eff~t~ Concentraion ~ CMCon Rate ~ Change ~ Flu~ity

a!
CMC Concentration
dt

z =[~,pt.~]

[,-']
-~ L-~-A 6.0 X 10-* 3 . 8 X 10-4 2.0 X 10-*

1.0 0.5
0.25

6 X 10-* 19 X 10-* 31 X 10.4

1.80 X 104 1.25 X 10"-2 0 . 8 4 X 10~

* Formula of Ingelman. Formula of Hultin: C ~, -- (concentration of s u b s t r a t u m ) 3

the incubated samples determined from the curve. In the fluidity method, the d of the incubated filtrate was compared directly with that of the original,

dt

a comparison previously shown to be valid for enzyme concentrations of the magnitude employed (Fig. 1). Seven filtrates were subjected to the heat treatment. The results obtained by the two methods are in fairly good agreement. In four cases, the amounts of inactivation obtained by the two methods are nearly identical. For the other three filtrates, the results obtained by measuring fluidity change are in line with the data obtained for the other filtrates. As indicated in the preceding section, the presence or absence of the substratum may be an important factor in enzyme inactivation. A similar situation appears in comparing the heat inactivation results with the temperatureactivity data. It will be recalled (Fig. 3) that the filtrate of Myrothecium verrucaria was less active at 60C. than at 50C., but that of Actinomyces sp. was more active at the higher temperature. Examination of the results of inactivation at 70C. (Table II) shows an entirely different picture, the Actino-

614

ENZYMATIC H Y D R O L Y S I S O F S O L U B L E C E L L U L O S E D E R I V A T I V E S

IOO -X

BO

60

4o

Zo

O.G

O~

I.O

1.2

DEGREE OF ~UBSTITUTION
FIG. 5. Effect o~ D I
Filtrates: From Aspergillus fumigatus QM 45h grown on cotton d u e l Conditions: CMC Samples: 50T DS: 0.52
70M 90M 120M 0.89 0,99 1.2

H. S. LEVINSON AND E. T. REESE

615

Asp. fumigatus filtrote + GMG 2 hours at 5 0 G . .6 .8 1,0 OEGREE OF SUBSTITUTION

1.2

)n enzyme activity.
Fluidity test: CMC 1 per cent; citric acid-NaOH buffer at pH 5.0, /20; filtrate 1:100. Temperature of test 35C. Reducing sugar test: CMC 0.5 per cent; K-PO4 buffer at pit 5.0, ~/50; filtrate 1:10. Temperature of test 50C. Time 2 hours.

616

ENZYMATIC HYDROLYSIS OF S O L U B L E C E L L U L O S E D E R I V A T I V E S

myces sp. filtrate being the much more rapidly inactivated. The data appear to support the view that inactivation proceeds more rapidly in the presence of substratum but do not nullify the conclusions previously drawn (in the section on activity) because of the short time periods employed in that study.

TABLE II
Temperature Inactivation of Cx
Duration of inactivation period
rain.

Inactivation* of enzyme Reducing sugar assay

per cent

Organism

Filtrate No,

Fluidity change assay

per cent

Aspergillus fumigatus QM 45h

H 26 H 11 L1 AM-X E 613 C 13 K3

30 120 30 120 30 120 30 120 30 120 30 120 30 120

39 72 70 82 61 74 55 70 93 97 80 90 100 100

65 82 71 82 61 74 64 77 67 78 80 90 100 100

"

"

"

"

~yrothedum verruc, arla QM 460

"

"

"

""

~tspergillus sydow4 QM 31c Frichoderma viride QM 6a 4ctinomyces sp. QM B814

* Inactivation, 70C.; pH 5.2.

I I L Activity of Filtrates of Various Origins Compared by Reducing Sugar and Viscosimetric Methods

If the two methods of assay employed measure the same reaction (i.e. one enzyme), then comparisons of various filtrates should show the same relative orders of activity. When a large miscellaneous group of filtrates was compared, the agreement between methods was very rough. A comparison of filtrates within an experiment should show better agreement. I n Table III-1, the results are from filtrates of organisms grown on glycerol. Under these conditions, the Cx activity is very low and the fluidity figures are based on a 1 : 10 dilution of the enzyme. I t will be observed that both assay methods are in agreement

H. S. L]~VINSON AND E. T. REESE TABLE III

Comparative Measurements of Filtrate Activity

617

Organism

Medium for growth

Glucose* 50 C.
mg.lml.

dl-- $
~/sp dt

Experiment 1
A s p e r g~Uus l u c h u e n s i s Myrothecium verrucaria Asperg~llus fumigatus Aspergillus luchuensis A c t i n o m y c e s sp.

Glycerol; 18 d a y old "

"

0.28 0,09 0.01 0.01 0.00

3 6 . 4 X 10-* 8.4 " " 1.4 " " 0.0 " 0.2 "
" "

"
"

9 d a y old

Experiment 2
Aspergillus fumigatus
gc gc

Duck Sliver Duck

400C. 30C. 30C.

0.50 0.34 0.22 0.10

79 X 10-4 57 " " 33 " " 18 " "

Solka-floc 30C. Experiment 3

Aspergillus ustus QM A s p e r g i l l u s terreus Aspergillus ustus Aspergillus tamarii A s p e r g i l l u s ver~icolor A spergillus fumigatus A spergillus flavipes AspergiUus niger A spergiUus flavus A spergillus davatus A s p e r g i l l u s ochraceus AspergiUus unguis

29c 91c 892 75b 17d 6b 24a 458 4m 862 26b 8f

CMC50T
" "

9dayold " " "

" " "

0.56 0.31 0.43 0.38 0.31 0.22 0.27 0.21 0.28 0.11 0.12 0.00 0.02

16.3 X 10-* 11.2 " " 6.0 " " 5.5 " 5.1 " 4.2 " 2.8 " 2.8 " 2.6 " 0.7 " 0.6 " 0.06" 0.03" " "
" " "

"
" " " " "

"
" " " " "

"
" " " " "

"
" " " " "

" "
"

" " "

"

"
"

"
" " "

" "
" "

"
"

"
"

A s p e r g i l l u s e a r b o n a r i u s 331

* Incubation time = Experiment 1, 2 hours; Experiment 2, 1 hours; E x p e r i m e n t 3, 1 hour; C M C 50T. :~ Filtrate concentration for fluidity test: Experiment 1, 1:10; Experiment 2 a n d Experim e n t 3, 1:100. T w e n t y d a y old culture of this organism showed appreciable Cx activity. even with filtrates of such a low level of activity. In another comparative the same organism
(Aspergillus fumigatus)

test,

was grown on different cellulosic

618

ENZYMATIC HYDROLYSIS 0~" SOLUBLE CELLIVLOSE DERIVATIVES

60610"4

/1
4(

d!/~SP
dt

rl

H26

g
K3J

pH OF INACTIVATION

FIO. 6. Effect of

Filtrates:
Filtrate No. H 26 Organism Substrate Cotton duck Cotton duck CMC Incubation time days 26 10 10

K 3 L 1

Aspergil~usfumigaus QM 45h Acgnomyces QM B814 sp. Myrolhedum verrucaria QM 460

Conditions:

Inaaiva~iom 1 ml.~s/2 citric add-NaOH buffer added to 9 ml. cell-free filtrate. Incubated 24 hours; 50C.

materials and at different temperatures. Again the relationship was good (Table III-2). Those filtrates most active in changing the fluidity were also most active in producing reducing sugars. Indeed the relationships are almost quantitative. Thus, when grown on solka-floc (a purified wood cellulose) at 30C., the organ-

H. S. LEVINSON AND E. T. REESE

619

I on Cx stability.

Fluidity test: CMC 70M 1 per cent; citric acid-NaOH at pH 5.0, m/20 in reaction mixture. Incubated filtrate diluted to 1:10 concentration and 10 ml. used per 100 ml. of reaction mixture. Temperature 35C. Reducing sugar test: CMC 50T 0.5 per cent; citric acid-NaOH at pH 5.0, aM/20 in reaction mixture. 5 ml. incubated filtrate brought to pH 5.0 with 1 ml. am/2 citrate buffer. Diluted with 5 ml. H20. 1 ml. of readjusted diluted filtrate used with 9 ml. CMC-citrate mixture. Temperature 500C.

ism produces a filtrate having only 20 per cent of the reducing sugar activity of the filtrate from duck at 40C., and 22 per cent of the liquefying ability. In the third test, many different organisms were grown under identical conditions and the filtrates compared by both assay methods. The agreement be-

620

ENZYMATIC ttYI)ROLYSIS O1~ S O L U B L E C E L L U L O S E D E R I V A T I V E S

tween methods is much poorer here. It can be said only that the top third of the list contains one set of organisms, the middle third a second, and the bottom third the rest. That is, there is a rough sort of agreement but one is forced to the conclusion that enzyme Cx may vary with the organism producing it (Table 111-3).

IV. Changes in Degree of Polymerization and in Reducing Sugar on Emymatic Hydrolysis of CMC
As was indicated previously, the proportionality between [7] and DP is a valuable tool. In the experiments reported here, no attempt was made to determine actual molecular weights. Instead, a comparison was made between the DP and the production of reducing sugar at various stages during the course of hydrolysis of a CMC solution. The filtrates used were those from Myrothecium verrucaria, Actinomyces sp., and Aspergillus fumigatus. In each of these experiments, reducing sugar measurements and viscosity measurements were made on the same batch of reaction mixture. Intrinsic viscosity determinations were made at intervals from 0 minutes' incubation to 24 or 48 hours' incubation, with the intervals more frequent in the initial phases of hydrolysis. The nature of the curves obtained is illustrated in Fig. 7, in which the filtrate of Myrothecium verrucaria was used. The intrinsic viscosities after each period of incubation are indicated along the ordinate. In order to determine the length of time necessary to reduce the average DP (or [7]) to half, quarter, eighth, etc. of its original value, [7] was plotted against the logarithm of the incubation time in minutes (Fig. 8). The initial phases of the hydrolysis proceed at a rate which appears to be a function of the logarithm of the incubation time. The value for the final [7/] approaches the same end-point, regardless of the source of the filtrate, pointing to the similarity of the active constituents of the various filtrates. Thus after 24 hours, the Myrothecium verrucaria filtrate brings the [~/]down to 6 per cent of the initial value and the Asperigillusfumigatus illtrate brings the [~/] to 7 per cent of the original in 22 hours. The average DP of the CMC at this time may be estimated very roughly as one-sixteenth that of the starting material. Parallel to the fluidity measurements, reducing sugar determinations were made on 1 ml. aliquots of the same reaction mixture. The results indicate that the filtrates of d'+~eient organisms behave similarly although specific differences are found. Of particular interest is the relationship between [~/] and amount of reducing sugar at various times during the hydrolysis. These have been plotted (Fig. 9) for filtrates of Myrothecium ~errucaria and of Aspergillus fumigatus. The curve for Actinomyces sp. falls between the other two, and is not shown on the graph. The Aspergillus furaigatus curve shows greatest change in [~/] with

H. S. LEVINSON A N D E. T. R E E S E

621

0 Mimers

~4
~.0

12

5 Ho~s

.14

GG M
FIO. 7. Intrinsic viscosity of CMC after incubation with Myrothecium ~rucari(z filtrate. Filtrate: Myrothecium v~rucaria QM 460 grown on cotton duck, 9 to 18 days and filtrates combined (filtrate AM 3;:). Conditions: CMC 70M 1 per cent; citric acid-NaOH buffer at pH 5.0, ~/20; filtrate 1:10 in reaction mixture. Action of filtrate stopped at different times by addition of 20 ml. of 20 per cent NaOH to 80 ml. of reaction mixture. Further dilutions with 4 per cent NaOH. least increase in reducing groups. The data presented earlier (Fig. 1) show that filtrates of this organism have limited ability to produce reducing groups from CMC. On the other hand, the most active of all filtrates in reducing the visco-

622

ENZYMATICHYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES

slty of CMC are those of Aspergillus fumigatus. These facts are in complete agreement. The filtrates used above have been tested for/~-glucosidase (cellobiase) activity by incubation with salicin at 50C. and pH 5.1. The reducing sugar values

s'o
TtMt/~wo'rtrl FIG. 8. Intrinsic viscosity changes in CMC during enzymatic hydrolysis. Filtrate:
Filtrate

1oo

No.

Organism

Substrate Cotton duck Cotton duck I-Iydroxyethyl cellulose

AM X H 26 K 2

Myrotht~ium verru~aria QM 460 AspergiUusfumigatus QM 45h Aainomyces sp. QM B814

Incubation time days Combined, 9-18 26 10

Conditions: CMC 70M 1 per cent; citric acid-NaOH buffer at pH 5.0, M/20; filtrate 1 : 10 (except K 2, 1:20) in final mixture.
obtained were less than 0.01 mg./ml./hr., in our opinionnegligible amounts. A filtrate of an old culture (47 days) of AspergiJius l~huensis was found to contain an appreciable amount of fl-glucosidase activity (0.24 mg./ml./hr, of glucose). Since this filtrate also had relatively low Cx activity (compared to those previously used), it was thought that results obtained with it might further elucidate the picture. The resulting curve (reducing sugar vs. [7]) almost completely coincided with that of Myrothecium vermcaria (Fig. 9), even though the

H, S. LEVINSON AND E. T. KEESE

62.3

rates of the reaction were greatly different. The hydrolysis of CMC thus appears to be a function of Cx, and to be unaffected by the presence of ~-glucosiii i j

?J

1.0 ~ ~ .

~ . ~ c m s 5 t ~ a as 6uJcost (m/mD FIG. 9. Reducing sugar and intrinsic viscosity changes during enzymatic hydrolysis of CMC,
Filtrate:
Filtrate No. AM X Organism
Substrate

H 26

Myrothecium verrucaria Q M 460 Aspergilluafumigatus QM 45h

Cotton duck Cotton duck

Incubation time days Combined, 9-18 26

Conditions: CMC 70M 1 per cent; citric acid-NaOH buffer at pH 5.0, ~t/20;

filtrate 1 : 10 in final reaction mixture. Temperature 35C. Values taken between 2 minutes and 24 hours of incubation with filtrate. dase. Since all filtrates follow the same pattern but are not identical, it is believed that Cx represents a group of related enzymes attacking the 1,4-/~-gluco-

624

ENZYMATIC HYDROLYSIS O]~ SOLUBLE CELLULOSE DERIVATIVES

sidic linkage as found in cellulose, but differing slightly from each other in certain specific aspects of their actions.
DISCUSSION

The enzymatic hydrolysis of soluble cellulose derivatives can be followed by observing the change in fluidity of the substratum. It is only necessary to slow down the process to the point where measurements can be made to cover the early action. This has been done by dilution of filtrate, use of a suboptimal temperature (35C.), and selection of a CMC preparation of desirable viscosity (CMC 70M at 1.0 per cent) and uniformity. Under these conditions quantitative measurements have shown that the specific fluidity increases at a constant rate during the early part of the reaction (Fig. 4); i. e.: until a particular level in the degrading process is approached. As a result, the length of time during which the rate is constant is an inverse function of the magnitude of the rate. In actual practice, it has been found desirable to select a rate measurable within the 1st hour of hydrolysis. With the initial fluidity kept constant, comparison of rates is a measure of relative activity. Thus, relative activities for several filtrates have been obtained under different conditions of pH, temperature, and enzyme concentration. Numerical values might be obtained by assigning an arbitrary unit figure to the enzyme filtrate giving a rate value of say 100 10-4. The actual establishment of a unit has not yet been made. The fluidity method has several advantages over the reducing sugar method, the most important of these being the possibility of following the rate of change in DP of the CMC molecule by following changes in intrinsic viscosity. With the fluidity method, early changes in DP are observable, while with the reducing sugar method, only the final products are determined. Since lower concentrations of the filtrates are used in the fluidity tests, a lower concentration of interfering impurities is introduced. On the other hand, with the reducing sugar method, it is possible to examine approximately four times as many filtrates in the course of a day as with the fluidity method. However, we are primarily concerned not with the superiority of one method over the other, but rather with the ability of the fluidity method to supplement the reducing sugar method in giving an insight into the mechanism of cellulose degradation. Does the fluidity method measure the same reaction that is followed by the reducing sugar measurement? Carboxymethyl cellulose is something of an unknown material. In its preparation, linter cellulose is oxidized under alkaline conditions, following which it is treated with monochloracetic acid. It is known that the degree of polymerization (DP) during the process is considerably reduced, the product having a final DP of about 150. The degree of substitution can also be determined, and is known for the samples used (i.e. CMC 70M, DS 0.89; CMC 50T, DS 0.52). It may safely be stated that such forces as are ex-

H . S. L E V I N S O N AND E. T. R E E S E

625

erted in holding chains together in "crystalline" cellulose are absent in CMC solutions. The existence of other types of cross-linkages between cellulose chains has been postulated. There is, for example, the hemiacetal linkage proposed by Pacsu (13). However, such a linkage is more easily split than the 1,4-/~-glucosidic linkage connecting the anhydroglucose units. Since even the latter are broken down during the manufacture of CMC, it seems unlikely that the hemiacetal linkage persists. It should further be pointed out that cross-linkages are generally introduced to account for very large molecules and not for those of DP less than 250 as found in CMC. The interpretation of our data rests on the assumption that CMC in solution is a straight chain molecule, and that the only linkage acted upon is the 1,4-~-glucosidic linkage. Examination of the data presented in this paper indicates great similarity in the results obtained by the two methods (fluidity, reducing sugars). The effects of pH on activity (Fig. 2) and on inactivation (Fig. 6) ; the effects of temperature on activity (Fig. 3) and on inactivation (Table II); and the effects of degree of substitution on activity as determined by viscosimetric measurements closely parallel those obtained by the reducing sugar procedure. Still other results confirm this. Filtrates have been treated with carbon and with ground cellulose at various pH levels. The results of assaying the filtrates and eluates by the fluidity method are in excellent agreement with those obtained by the reducing sugar method. These are data based on comparison of like filtrates by dissimilar methods. The evidence substantiates the assumption of a single enzyme (Cx) acting upon a particular linkage (1,4-/~-glucosidic) found in the cellulose derivatives. On the other hand, differences have been found between the enzyme Cx produced by different organisms. Thus, variations in pH and temperature optima (Figs. 2, 3), in heat stability (Table II) and in other respects have been pointed out. (These variations, it is emphasized, are between organisms and not between methods of assay.) Such differences between enzymes from different sources have been previously reported. Proofs that the enzyme itself varies--and that the results are not due to inhibitors, etc. in the media--are less abundant. Studies on a-amylase, however, offer convincing evidence that variation in the enzyme does exist. Hanes and Cattle (5) combined measurements of reducing sugar with measurements of the iodine-starch color complex during hydrolysis. A plot of their results produced different curves for enzymes from different sources. Similarly, Miwa el al. (12) have compared the/$-glucosidases of various organisms with respect to their abilities to hydrolyze several different fl-glucosides. These results, too, demonstrate variation in enzyme from different organisms. Our own data (Fig. 9) are similar to those of Hanes and Cattle. When [7] (in place of the starch-iodine complex) is plotted against reducing sugars, different curves result for preparations derived from different organisms. In combination with other data presented above, it appears reasonable to conclude

626

ENZYMATIC ttYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES

that Cx, like a-amylase and ~-glucosidase, represents a family of enzymes similar in that they act on a common linkage but differing in certain specific properties. While we have been impressed with the essential similarities of the results obtained by the two methods of measurement, we have not been unaware of differences. As shown above, we attribute such differences to variation in an enzyme found within a class (Cx). It should be emphasized however, that the differences may be interpreted on an hypothesis of two e n z y m e s m n e effecting a rapid change in viscosity, the second forming reducing sugars. The question arises as to the place of the enzyme Cx in cdhdose hydrolysis. It should be clearly understood (a) that Cx is naturally occurring, i.e. it is found in filtrates from cultures of organisms growing on cellulose and (b) that these filtrates have the ability to hydrolyze cellulose, alkali cellulose, and cellulose dextrins as well as the cellulose derivatives used in this study. We are not, then, dealing with enzymes adapted to a particular, and unnatural substratum. The currently accepted explanation of cellulose hydrolysis is the early view that
Cellulose cellulase ~ cellobiose cellobiase ) glucose

Proof of the validity of the reaction rests on the establishment of cellobiose as an intermediate. Early workers did present evidence that seemed to indicate that cellobiose was one of the products of hydrolysis of cellulose. Pringsheim (14, cited by Waksman), succeeded in demonstrating both cellobiose and glu~ cose in the hydrolysis products of thermophilic bacteria acting on cellulose. Cell-free filtrates of the organisms were not used, however, and the possibility of intracellular enzymes being released on lysis of the cells exists. In fact, in the experiments cited by Waksman, bacterial action was brought to a halt by addition of an antiseptic at 65C. which may have enhanced lysis of the organisms. On the other hand, Kalnins (9, cited by Norman and Fuller) checked the fermentation of cellulose by mesophilic organisms at 37C. and could not detect cellobiose. In no case have we been able to discover substantial proof that cellfree filtrates of aerobic mesophilic organisms degrade cellulose to glucose mainly through a cellobiose intermediate. A deviation from the above picture appears in the work of Grassmann et al. (4). Two enzymes were separated from a dialyzed enzyme solution. The first, "cellulase," was capable of hydrolyzing cellulose in chains having as few as six anhydroglucose units. The second, cellobiase (~-glucosidase) hydrolyzed chains of two to six units in length and/~-methylglucoside, but had very little action on the larger molecules. The intermediate, cellobiose, was thus replaced by a mixture of soluble fragments of the cellulose chains.
Cellulose 1,4 B-polysaccharase "cellulase" --~ 1,4-B-oligosaccharase

soluble fragments
n=2--6

(C~Ito05)~

ceUobiase

> glucose

It. S. L E V I N S O N AND E. T. R E E S E

627

This series closely resembles that recently proposed by us (15) c~ Cellulose ...... --~ (C6H1005)~ straight chain fragments Cx ) glucose

Cellulolyficorganisms Non-cellulolyticorganisms

I I

in which it was shown that many non-cellulolytic organisms can hydrolyze soluble cellulose derivatives. A ceUulolytic organism in the presence of cellulose produces an enzyme, C1, which alters the native cellulose in some unknown manner, making it available to the hydrolytic enzyme Cx. It may be that the preliminary action is of such a nature that individual chains are separated sufficiently for Cx to be able to act upon them. Since Cx is capable of hydrolyzing long chains, it is likely that the separate chains are long and so insoluble, but Cx appears also to be able to hydrolyze those fragments that are soluble (i.e. where n is less than ten). The enzyme Cx splits the chain at random, forming progressively smaller fragments. The size of the particle capable of entering the microbial cell is unknown, but glucose and cellobiose, and probably also larger soluble molecules, pass in. There is thus no necessity for an extracellular cellobiase, and indeed we have found none (unreported). The authors are grateful to Dr. H. M. Spurlin and the staff at the Hercules Powder Company for their many valuable suggestions, and for a generous supply of CMC of known properties. We also wish to acknowledge the valuable guidance of Dr. R. K: Ladisch, Dr. G. R. Mandels, and Dr. R. G. H. Siu of our own laboratories.
SUMMARY

Observation of changes in fluidity is presented as a method for following the enzymatic hydrolysis of soluble cellulose derivatives. The activity of different ceU-free enzyme preparations may be compared by this method, providing certain precautions are observed. In general, results obtained by use of the fluidity method are similar to those obtained using the reducing sugar technique, indicating that the same enzyme system is measured by the two methods. Changes in the DP of the substratum may be followed within certain limits of molecular size. Results indicate that a random splitting of CMC occurs during enzymatic hydrolysis, with a concomitant decrease in intrinsic viscosity and an increase in reducing sugars. Certain inadequacies of the cellulose-cellobiose-glucose theory, together with more recent findings, have led to the postulation of an alternate explanation of the mechanism of cellulose hydrolysis.

628

ENZYMATIC HYDROLYSIS OF SOLUBLE CELLULOSE DERIVATIVES

REFERENCES

1. Northrop, J. H., and Hussey, R. O., 1923, J. Gen. Physiol., 5, 353. 2. Bamann, E., and Myrback, K., 1941, Methoden der Fermentforschung, 9., 1842. 3. Brown, C. J., and Houghton, A. A., 1941, Chemical and physical properties of carboxymethyl cellulose and its salts, Soc. Chem. Ind., Tr., 60, 254. 4. Grassmann, W., Zechmeister, L., Toth, G., and Stadler, R., 1933, Uber den enzymatischen Abbau der Cellulose und ihrer Spaltprodukte, Ann. Chem., Berlin, 503, 167. 5. Hanes, C. S., and Cattle, M., 1938, Starch-iodine coloration as an index of differential degradation by the amylases, Proe. Roy. Soc. London, Series B, :[25, 387. 6. Hultin, E., 1948, Viscosimetric assay of hyaluronidase, S~ensk Kern. Tidskr., 60, 131. 7. Ingelman, B., and Malmgren, H., 1947, Enzymatic breakdown of polymetaphosphate, Acta chem. scand., 1, 422. 8. Ingelman, B., and Malmgren, H., 1949, Enzymatic breakdown of polymetaphosphate, Aaa chem. scand., 8, 157. 9. Kahains, A., 1930, Acta Univ. Latviensis Lauksaimniecibas Fakult~es, 11, series 1,221, cited by Norman, A. G., and Fuller, W. H., 1942, Cellulose decomposition, Advances Enyzmol., 2,239. 10. Letzig, E., 1942, Enzymatische Nachweis yon Binde--- und Verdickungsmitteln, Z. Untersuch. Lebensm., 84, 289. 11. Letzig, E., 1943, Beitrag zur Frage der Verdaulichkeit und Unschadlichkeit wasserl6slicher Cellulosederivate, Z. Untersuch. Lebensm., 85, 401. 12. Miwa, T., Cheng, C., Fujisaki, M., and Toishi, A., 1937, Acta Phytoddm. Japan, 10, 155, cited by Pigman, W. W., 1944, Specificity, classification, and mechanism of action of the glycosidases, Advances En~.ymol., 3, 41. 13. Pacsu, E., 1947, Molecular structure of cellulose and starch, J. Polymer Sc., 2, 565. 14. Pringsheim, H., 1912, Uber der fermentativen Abbau der Zellulose, Z. physiol. Chem., ?8, 266, cited by Waksman, S., Principles of Soil Microbiology, Baltimore, The Williams & Wilkins Co., 1932. 15. Reese, E. T., Siu, R. G. H., and Levinson, H. S., 1950, Biological degradation of soluble cellulose derivatives, and its relationship to the mechanism of cellulose hydrolysis, J. Bact., in press. 16. Siu, R. G. H., 1950, Microbial action on cellulose, unpublished manuscript. 17. Sumner, J. B., and Somers, G. F., 1944, Laboratory Experiments in Biological Chemistry, New York, Academic Press, Inc. 18. Ziese, W., 1931, Uber die Einwirkung yon Fermenten des Magensaftes yon Helix pomatia and solcher des Gerstenmaldes auf Celluloseglykol~ther, Z. physiol. Chem., 203, 87.