Lysosomal Hydrolase Secretion by Tetrahymena: A Comparison of Several lntralysosomal Enzymes with the lsoenzymes Released into the Medium J. J. BLUM Department of Physiology and Pharmacology, Duke University Medical Center, Durham, North Carolina 27710

ABSTRACT Tetrahymena pyriformis were grown i n proteose-peptone medium and then washed and incubated i n a dilute salt solution for one hour. The cells were then discarded and the lysosomal hydrolases that had been secreted were subjected to DEAE cellulose column chromatography. At least three isoenzymes of acid phosphatase, three of acid protease, and two of p-N-acetylhexoseaminidase were found, as well as single peaks of a-mannosidase, pgalactosidase, and p-fucosidase. The latter two activities were not resolved by the DEAE column and could not be separated i n a second chromatographic step on CM-cellulose. Cells were also grown under identical conditions and homogenized in 0.25 M sucrose in order to allow comparison of some of the intracellular lysosomal hydrolases with their secreted counterparts. Two lysosomal populations were resolved by sucrose density gradient sedimentation, a heavy lysosomal fraction, centered at a density of about 1.25 gmlcm3, and a light lysosomal fraction, centered at a density of about 1.16 gm/cm? These two populations differed i n that the light lysosomes did not appear to contain significant amounts of pfucosidase, p-galactosidase, or acid protease, whereas all six of the hydrolase activities studied were present in the heavy lysosomes. The light lysosomal peak occurred in cells grown to transition phase, but was markedly reduced in cells from cultures grown to stationary phase. In addition to these two fractions a third very light particle, containing only a-mannosidase activity, was detected just inside the gradient. Measurements were made of the effect of heat (10 minutes at 66") and of a change i n pH from 4.5 (standard assay condition) to 6.0 on the three acid phosphatases and two p-N-acetylhexoseaminidase isoenzymes resolved by DEAE column chromatography of the secreted hydrolases and on these hydrolyases in the heavy and light lysosomal fractions on the sucrose gradient. Use of the thermostability and pH criteria permitted computation of the expected properties of the intralysosomal acid phosphatase and hexoseaminidase activities if these consisted of the respective isoenzymes in the proportions secreted. It was found that neither the intralysosomal acid phosphatase nor the intralysosomal hexoseaminidase had the properties expected if they consisted of the secreted mixture of the respective isoenzymes, indicating that modification of some of these isoenzymes may have occurred during the 1-hour starvation period or after secretion.

Since the discovery (Muller, '70) that Tetrahymena release large quantities of several lysosomal hydrolases into the medium, much has been learned about several factors controlling the rate of secretion. With increasing culture age, for example, the amounts of acid phosphatase, a-mannosidase, p-fucosidase, and p-galacJ

CELL P n ~ s r o ~89. . . 457-472.

tosidase secreted in one hour all increase while the amounts of RNAase and a-glucosidase decrease (Rothstein and Blum, '74a; Blum, '75). It was also noted that growth of cultures in proteose-peptone medium supplemented with glucose (among other Received Jan 8, '76 Accepted M a r 2. '76

457

458

J. J. BLUM

metabolites) caused a marked decrease in the amounts of six lysosomal hydrolases secreted during a subsequent 1-hour incubation i n dilute salt solution as compared to cells grown i n unsupplemented medium (Blum, '75). Evidence has also accumulated that the lysosomal population of Tetrahymena is heterogeneous (Blum and Rothstein, '75) and that hydrolases from the particles which band at high densities are preferentially secreted into the medium (Muller, '72; Rothstein and Blum, '74b). These observations raise many questions as to the factors controlling the rate of release of different hydrolases from the various classes of lysosomes, questions which are of necessity ignored when only the secretion of one or two hydrolases is examined. Furthermore, earlier work suggesting that two or more protease activities were released into the medium (Dickie and Liener, '62; Blum, '75) also raises the question as to whether any of the other hydrolases are secreted i n more than one form, and, if so, what are the physiological factors which control the rate of secretion of one isoenzyme relative to its other form(s). In the present paper we have studied six acid hydrolases secreted by Tetrahymena during a 1-hour incubation and subjected them to chromatography on DEAE-cellulose columns. Since very little protein is released into the medium, even a single column step yields considerable resolution of the enzyme activities. Our results show that at least three acid phosphatase isoenzymes, two p-N-acetylhexoseaminidases, and three or more acid proteases are secreted whereas a-mannosidase, p-galactosidase, and p-fucosidase appear to be secreted largely as single enzyme species. Preliminary characterization of the two forms of p-Nacetylhexoseaminidase and of some of the acid phosphatase isoenzymes has been achieved by comparing their heat stabilities and the ratio of activities at pH 6.0 to those at pH 4.5. These studies were performed on transition-phase cells (grown with or without glucose supplementation) and on cells in stationary phase. Experiments were also performed on cells grown under identical conditions but homogenized prior to the 1-hour incubation i n dilute salt solution. Homogenates of the cells which had not been permitted to secrete

were layered on sucrose density gradients and the distribution of the acid hydrolases in the gradient examined, and the properties of some of these intracellular hydrolases were compared with the properties of the secreted hydrolases. MATERIALS A N D METHODS

Growth conditions and harvesting Tetrahymena pyriformis, strain HSM, were grown at 25" i n a medium of 1% proteose peptone and 0.05 % liver extract i n 20 mM potassium phosphate adjusted to pH 6.5 with NaOH. Cells were grown in a gyrotary shaker in 500 ml Erlenmeyer flasks equipped with Morton closure tops and containing 130 ml of culture. Cells were grown for either 17-18 hours (referred to as transitional cultures), 41-42 hours (referred to as stationary cultures), or for 17-18 hours in the presence of an initial concentration of 15.4 mM glucose (referred to as transitional cultures glucose). Cells were counted with a Coulter counter (Coulter Electronics, Inc., Hialeah, Florida). For experiments in which the properties of secreted hydrolases were to be studied, cells from four cultures were centrifuged at room temperature for three minutes at 200 X g , the pellet resuspended in dilute Wagner's solution (Rothstein and Blum, '73), and recentrifuged for three minutes at 200 X g . The cell pellet was again resuspended in the dilute Wagner's solution and centrifuged for two minutes, and the washed cells were finally resuspended i n r u 8 2 ml of 10 mM NaH2P04,pH 7.0, containing 500 pg chymostatin (Blum, '75). The cell density was about 2 X 1 0 6 cells/ml. Twenty milliliter portions were placed into each of four 1-liter capacity Erlenmeyer flasks and incubated at 25" without shaking for one hour. When desired, a second set of four cultures was processed in the same way immediately after the first set. At the end of the 1-hour incubation, the cultures were centrifuged for three minutes at 200 X g at O " , and the pooled supernatant was again centrifuged at 0" to remove any remaining cells. The combined cell-free supernatants, containing the secreted lysosomal hydrolases, were then concentrated about 10-fold by ultrafiltration at 0" through a UM-10

+

LYSOSOMAL HYDROLASE RELEASE

membrane in a n Amicon ultrafiltration apparatus using Nz gas at -30 psi. The concentrated hydrolase secretion was dialyzed overnight at 4 " against a total of l liter of 0.01 M NazHP04, pH 8.0. Aliquots were taken for enzyme assay and the remainder placed on a DEAE column (see below). For experiments in which the distribution and properties of the lysosomal hydrolases were to be measured in cell homogenates, single cultures were collected as described in detail elsewhere (Rothstein and Blum, '74b). After homogenization i n 4.5 ml of 0.25 M sucrose, the volume was adjusted to 2 5 4 0 ml of 0.25 M sucrose (corresponding to 2 4 . lo6cells/ml) and 1 ml was placed on 21 ml of a linear sucrose density gradient (17.5 % to 58 % , w/w, resting on a 2-ml cushion of 65% (w/w) sucrose). Gradients were centrifuged for four hours at 24,000 RPM i n the SW 25 rotor of a Spinco Model L ultracentrifuge at 0 " rotor temperature. Twenty fractions of 50 drops (w1.2 ml) were collected from the bottom of each gradient and kept on ice until samples were taken for enzyme assay. Column chromatography About 100 gm of DEAE 52, preswollen, was prepared by washing with 400 ml 0.1 N HC1, then distilled water, then 300 ml of 0.1 N NaOH, followed by copious rinsing with distilled water. The resin was then washed with 0.2 M NaHZPO4,pH 7.6, and then i n 0.01 M NaHzP04,pH 7.7. A column 44 cm high by 2.5 cm diameter was poured and washed with 0.01 M NaHZPO4, pH 7.7 buffer until the effluent was also pH 7.7. The dialyzed enzyme mixture, containing about 3 mg of protein, was then layered onto the column and washed i n with the same volume of pH 7.7 buffer. Proteins were then eluted with the output from a pair of cylindrical reservoirs with about 225 ml of pH 7.7 buffer in the proximal reservoir (which was well stirred via a Magnetic stirrer) and about 220 ml of pH 7.7 buffer containing 0.5 M NaCl in the distal reservoir. This would have provided a linear gradient of NaCl except that there was a "dead space" of about 50 ml above the resin which served to make a very shallow initial rise in NaCl concentration before there was a sigmoidal

459

increase to the final level, as shown i n figure 1. After about 425 ml of eluant had been collected, 100 ml of 0.01 M NaH2P04, 0.5 M NaC1, pH 7.7, was added to the distal reservoir. Fractions of about 4.5 ml were collected at a flow rate of approximately 1 mllmin. Fractions were stored frozen (1-3 days) until assayed. Chloride was measured titrimetrically using a Buchler chloridimeter. Protein was not measured after a n initial trial which showed that most of the protein came off in the fractions near and just after the initial peak of acid phosphatase activity and even here the maximum protein concentration was less than 250 pg/ml. Assay procedures All assays with p-nitrophenylated substrates were performed at 25" as described earlier (Blum, '75). Activities are expressed as nmoles hydrolyzed/hr.ml of fraction. For the acid protease assays aliquots of the fractions from the DEAE column were incubated with 0.5 ml hemoglobin substrate (total volume, 1.0 ml) at 25" for up to 18 hours. The reaction was stopped by adding 0.5 ml ice cold 5% (wlv) trichloroacetic acid and the tubes centrifuged at 0" for ten minutes at 12,000 X g. Portions of the clear supernatant were taken, neutralized with NaOH, and assayed for peptides by the method of Benson and Hare ('75). Fluorescence was measured in a Model A 3 Farrand fluorometer, using a Corning 7-51 filter i n the activating light path and a Corning 3-73 filter i n the fluorescent light path. For assays of fractions from sucrose gradients blank samples were also run in which identical aliquots were incubated with hemoglobin and trichloroacetic acid for the same times (618 hours) as the test samples, which were incubated with the hemoglobin substrate alone. These blanks served to correct for any acid-soluble peptides present i n any of the fractions, so that the difference in fluorescence between any tube and the corresponding control was due only to the protease activity. Standards containing known amounts of bovine serum albumin (BSA) were run under conditions identical to those in the assay of material from either the DEAE columns or from the sucrose gradients; protease activity is expressed as k g BSA equivalents/hr,ml fraction.

460

J. J. BLUM

Reagents p-Chloromercuribenzoate and p-nitrophenylated substrates were purchased from the Sigma Chemical Co., St. Louis, Missouri; DE52, microgranular, from Whatman Biochemicals, Ltd., England; O-phthalaldehyde from the Eastman Chemical Co., Rochester, New York. D-glucono-1,5-lactone, L mannono-l,4-lactone, and a-glucoheptonic lactone were purchased from Pfanstiehl Chemical Co. All other reagents were of the highest purity commercially available. RESULTS

A typical chromatogram of the hydrolases secreted by cells from a transition phase culture is shown in figure 1. Three such experiments were done with hydrolases released from cells grown for two days (stationary cells; see MATERIALS AND METHODS) and four experiments with cells grown for about 18 hours (transition cells). Three experiments were also performed with the hydrolases secreted by transition cells grown with glucose but inconsistent results were obtained and the data are not reported here. Since we are unable to establish any clear differences i n the pattern of activity released by transition as compared to stationary cells, we have combined the data for these seven experiments. Table 1 shows the percent recovery of the various hydrolases assayed except for the proteases, which were assayed in only a few experiments and TABLE

1

Average r e c o v e r y of h y d r o l a s e a c t i v i t i e s from D E A E c o l u m n s Enzyme

Acid phosphatase peak I all other p-N-acetylhexoseaminidase peak A peak B

r,; recovery

44211 55 2 2 4 99 44k13 54+19

98

p-galactosidase P-fucosidase a-mannosidase

66i14 23 k 17 16k8

Data have been pooled from three experiments o n stationary cells plus four experiments on transition cells, i . e . , 11 = 7 In this and i n tables 2 and 3 . the m e a n k one standard deviation

which in some cases was much larger than l o o % , perhaps because of incomplete removal of chymostatin during the dialysis followed by complete removal during chromatography. I. Elution profile and some properties of lysosornal hydrolases secreted by Tetrahymena Acid protease Experiments on the effect of proteolytic inhibitors on the protease activity secreted by Tetrahymena indicated that there were at least two types of protease released, in agreement with the earlier work of Dickie and Liener ('62) and of Blum ('75). Figure 1 shows that there were at least three and probably five acid proteases resolved by DEAE chromatography of the released hydrolases. Since we have not characterized these acid proteases as yet, we have not assigned numbers to them. It is of interest, however, that Levy et al. ('76) have recently studied the intracellular proteases of Tetrahymena and find four or five peaks of acid protease activity. It appears likely, therefore, that each of the intracellular proteases is secreted, but further work is necessary to correlate the released proteases with their intracellular counterparts. p-fucosidase and p-galactosidase p-fucosidase and p-galactosidase always come off the DEAE column at exactl y the same position. In one experiment this peak was concentrated by filtration, dialyzed against 0.01 M phosphate, pH 6.0, and applied to a Sephadex CM 50 column equilibrated with this buffer. Proteins were eluted by increasing NaCl i n this buffer; the p-galactosidase and pfucosidase activities were eluted at the same position (data not shown). We have earlier pointed out (Blum, '75) that the amount of p-fucosidase secreted doubles in cells grown for two days as compared to cells grown for one day, whereas there was hardly any change in the amount of p-galactosidase activity secreted. Table 1 shows that about 6 6 % of the p-galactosidase activity put on the DEAE column was recovered, whereas only about 23% of the p-fucosidase activity was recovered. Since the relative amounts of these two activities secreted appear to vary inde-

461

LYSOSOMAL HYDROLASE RELEASE

pendently with culture age and the two activities have different stabilities during chromatography, it seems probable that these glycosidases are closely related but not identical. It should also be noted that i n this experiment and i n several others there appears to be a small broad peak of both p-fucosidase and p-galactosidase activity which was eluted well after the main peak (fig. 1). Since the amounts of each of these activities relative to the main peak were variable from experiment to experiment, we do not know whether these activities represent small amounts of isoenzymes of p-galactosidase and pfucosidase or are artifacts caused, for example, by protease activity that had not been inhibited by the chymostatin.

a-mannosidase a-mannosidase was eluted at about 0.25 M NaC1, usually as a large asymmetric peak [followed by a broad peak of much lower activity (fig. I), which may represent an isoenzyme]. The recovery of LYmannosidase activity was generally the lowest of all the hydrolases studied (table 1). p-N-acetylhexoseaminidase There were two peaks of p-N-acetylhexoseaminidase activity (fig. 1). The average recovery of this hydrolase was 98% of the amount put on the column; 44% of the activity was in the first peak, peak A, and 5 4 % was in peak B (table 1). Because of the large variability in the amounts

I

a, m

-2000

I0.000c

Acid phosphatase

a! m

-N-acetyl hexoseaminidase

2 8000 S

a

b 6000

c

-

1600 E

-

1200

0 a,

f

-800

2 4000

0 a,

2a 0 u

U

x a,

400

2000

L

c

c &

I 0

a,

E

a,

a

2

g

20

40

60

80

100

I20

I40

I60

I00 0

FRACTION NUMBER Fig. 1 Resolution of some hydrolase activities secreted by Tetrahymena. Eight 500 ml Erlenmeyer flasks containing 120 ml each of a culture at a cell density of 50,800 cellslml were grown for 18 hours to a final cell density of about 583,000 cells/ml. The cells were then collected, washed, and allowed to secrete for one hour as described i n MATERIALS A N D METHODS. The cell free supernatant (141 ml) was concentrated to 13.5 ml, dialyzed against 0.01 M Na2HP04,pH 8.0, and 11 ml of the concentrated dialyzed material, representing the hydrolase secretion i n one hour by 0.22 x 109 cells, were applied to the DEAE column as described in MATERIALS A N D METHODS. The activity of each enzyme i n the fluid applied to the column and the % activity recovered were, respectively: acid phosphatase, 46.9 bmoleslhr.rn1, 9 9 % ; p N-acetylhexoseaminidase, 7.27 pmoles/hr.ml, 105 5% ; p-galactosidase, 10.2 pmoles/hr,ml, 7 6 % ; p-fucosidase, 0.457 fimoles/hr,ml, 5 7 9 ; a-mannosldase, 1.94 pmoles/hr.ml, 21 % . The NaCl concentration is indicated by the dotted line in the middle panel.

p

462

J . J. BLUM

recovered, the difference between the hexoseaminidases A and B when assayed two peaks is not significant, and it is at pH 6.0 were 15% and 60%, respectivepossible that they could be released i n ly, of their activities at pH 4.5. Hexoseequal proportions. Hexoseaminidase B aminidase A was also more sensitive to is eluted close to the peak ofa-mannosidase, heating than the B isoenzyme; the A isobut the peak tubes were usually separated enzyme lost over three-fourths of its acby at least two fractions. Because of the tivity after heating 66" for ten mininterest in hexoseaminidase isoenzymes utes, whereas the B iosenzyme retained in relation to human lysosomal disease most of its activity under these conditions and in order to compare the secreted iso- (table 2). enzymes with the intracellular hexoseaminidase activity, some preliminary char- Acid phosphatase Peak I of acid phosphatase did not abacterization of these isoenzymes was undertaken. Kanfer and Spielvogel ('73) sorb to DEAE at pH 7.7 and was eluted at reported that several lactones inhibited the column volume (fig. 1). It was always both a highly purified p-N-acetylhexose- the largest of the acid phosphatase peaks, aminidase obtained from commercial bo- accounting for about 44% of the total acid vine serum albumin and the hexoseamini- phosphatase activity (table 1). It can be dase activity in rat liver lysosomes, which seen that peaks I1 and I11 were not well contain isoenzymes A and B. Hexoseamini- resolved under these conditions, and in dases A and B secreted by T e t r a h y m e n a many cases (as in fig. 1) there were indica(which may not, of course, correspond i n tions of other acid phosphatase isoenzymes any way to the A and B isoenzymes from as well. Acid phosphatase I loses virtually mammalian tissue) were not inhibited by all of its activity when heated for ten D-glucoso-l,5-lactone, L mannono-1,4- minutes at 66", but loses only about half lactone, or a-glucoheptonic lactone even of its pH 4.5 activity when assayed at when the concentration of lactone was 5 pH 6.0 or in the presence of 0.2 mM pmM, i.e. 2.5 times the concentration of chloromercuribenzoate (table 2). These p-nitrophenyl-glucoseaminide (data not characteristics are sufficient to differshown). At least in this property, there- entiate it from either isoenzyme I1 or 111. fore, the protozoan hexoseaminidases dif- Because the resolution of acid phosphatase fer from their mammalian counterparts. I1 from acid phosphatase I11 was poor, it is Two experiments were done in which 0.2 likely that the properties shown in table 2 mM p-chloromercuribenzoate was added for these isoenzymes refer to varying mixto each isoenzyme and the activity com- tures of the two forms. Despite the unpared to that in the absence of this SH- certainty arising from this cross-contaminareactive reagent. No inhibition was ob- tion, it seems likely that acid phosphatase served in either experiment. The two I1 has lower activity at pH 6.0 as compared isoenzymes differ appreciably in their sen- to pH 4.5 than does acid phosphatase 111, sitivity to changes in pH and in their and it certainly is more thermostable than sensitivity to heating. The activities of is acid phosphatase I11 (table 2). The data TABLE 2

S e n s i t i v i t y of i s o e n z y m e s of acid p h o s p h a t a s e a n d g - N - a c e ty lh e x o s e a m in id a s e to h e a t , p H , a n d p-chloromercuribenzoate J3-N-acetylhexoseaminidase

Acid phosphatase

Treatment

A

I

B

I1

111

23.5 f 3.9(6) 67.7 &8.7(6) 86.0 ? 6.3(6)

40.8 % 16.4(6) 7 . 7 %5.7(6) 72.0 f 18.3(5)

c/r, activity r e m a i n i n g

p H 6.0 10 min at 66" 0.2 m M PCMB

15.5 f 2.2(7) 23.0 f 6.4(7) 99 (n = 2)

59.4 f 9.0(7) 88.4f 4.4(7) 99 (n = 2)

57.3 f 17.5(7) 0.93 +-0.15(7) 49.1 f 9.1(7)

Numbers in parentheses are the number of experiments used out of a total of four o n transition cells and three on stationary cells. In each case 100% activity refers to the activity measured at 25' and p H 4.5, i.e., the standard assay conditions as described i n MATERIALS AND METHODS. PCMB is a n abbreviation for p-chloromercuribenzoate.

LYSOSOMAL HYDROLASE RELEASE

do not permit one to decide whether acid phosphatase I1 is significantly less inhibited by p-chloromercuribenzoate.

463

Acid protease This hydrolase is found mainly i n the lower portion of the gradient where the mitochondria and heavy lysosomes band 11. Distribution of particle-bound (Rothstein and Blum, ’74b). There may be hydrolases on sucrose gradients some acid protease activity i n the lower density region (fractions 11-13), but this Several studies (see Blum and Rothstein, is at best a very small peak. This does not ’75, for review) indicate that there is con- agree with Miiller’s (’71) finding that the siderable heterogeneity of the lysosomal low density particles from homogenates particles of Tetrahymena as well as in of Tetrahymena pyriformis, syngen 1, matmany other species (Davies, ’75). Muller ing type 11, were particularly rich in pro(‘72) showed that there were at least two tease activity. Whether this difference is populations, and that the main source of due to the difference in strains studied or the secreted hydrolases appeared to be to the fact that Muller’s experiments the population which banded at a high were done with cells grown in a synthetic density in sucrose gradients, and similar medium cannot be decided at present, but it observations were reported by Rothstein is clear that the distribution of protease and Blum (‘74b). Lloyd et al. (‘71) showed activities in lysosomal subpopulations of that the distribution of acid phosphatase, Tetrahymena must depend on one of these p-N-acetylglucoseaminidase, and acid factors. It should be noted that there was DNAase i n zonal gradients varied with always more soluble acid protease activity culture age and with starvation, each en- (at the top of the gradient in figs. 2-4) zyme according to a separate pattern. It than there was i n the lysosomes. It is not was therefore of interest to study the dis- known whether this arises because of tribution of the acid hydrolases studied damage to the lysosomes during homogenin the present work i n homogenates of ization or if there is an appreciable level Tetrahymena which were subjected to of soluble acid protease activity i n the cysucrose density gradient sedimentation, tosol. Because of the relatively small numespecially since there have been no pre- ber of experiments, it cannot be decided vious studies of the distribution of p- whether there were any significant diffucosidase, p-galactosidase, or or-mannos- ferences in the distribution of acid proidase within the lysosomal population of tease activity as a function of the growth these cells. Furthermore, knowledge of conditions investigated. some of the properties of the secreted isoenzymes permits one to ask whether there are p-fucosidase and p-galactosidase differences between the intracellular conThese hydrolase activities were the lowtent of the isoenzymes and the proportion est in amount of the six hydrolases assayed, secreted. Cells were grown exactly as for the and were approximately equal, as expected studies on hydrolase secretion, but were from earlier work (Blum, ’75). Most of washed i n ice-cold dilute Wagner’s solu- the activity appeared in the lower portion tion and gently homogenized as described of the gradient (near fraction 6) and there in MATERIALS AND METHODS. A n aliquot was no indication of any peak i n the lighter of the homogenate was then layered onto portion of the gradient in any of the 11 a continuous sucrose gradient as described such gradients prepared with the sole exin MATERIALS AND METHODS and fractions ception of the gradient shown in figure 2, of 0.1 1.2 ml were collected and assayed. where a trace of such a peak may exist. Four such experiments were conducted In most of the gradients there appeared with transition cells and with stationary to be a small peak of activity in a very cells and three experiments with cells grown dense particle (fraction 3) but this did to transition phase with glucose. Typical not occur in all gradients. gradients for transition cells, stationary cells, and transition cells grown with a-mannosidase glucose are shown in figures 2, 3, and 4, a-Mannosidase activity occurred in three locations within the gradient. One peak respectively .

464

J. J. BLUM

A

Acid phosphatase

I800 aJ

Y)

2 1500L

a

g 1200-

r a

900u

a

600 -

300t 0

o(-

rnannosidose

F R A C T I O N NUMBER

Fig 2 Distribution of six hydrolase activities i n sucrose gradients of homogenates from transition phase cultures Cells were grown for 19 hours from a n initial density of 54,400 cellsiml to a final density of 555.000 cells/ml, and then collected, homogenized, and layered onto a sucrose density gradient a s described in MATERIALS AND METHODS Acid protease activity is given as p g of bovine serum albumin equivalents hydrolyzedhr ml of fraction, all other activities are expressed a s nmoleslhr ml of fraction Percent recoveries were acld protease, 8 5 % , p-fucosidase, l O O @ ’ r , p galactosidase, 8 8 % , a rnannosidase, 1 3 9 4 , p - N acetylhexose aminidase, 120%, acid phosphatase, 115% The approximate density of each fraction is indicated by the abscissa at the top of this graph

of activity was always localized with the main mitochondrial and lysosomal peak at about fraction 6 or 7. A second peak at about fractions 11-13 was always present in transition cells, whether grown in the presence or absence of glucose. In cells grown to stationary phase, however, this second peak of a-mannosidase activity was either entirely absent or, as shown in figure 3 , reduced to a shoulder. A third peak

always was found just inside the gradient (fraction 17). This peak was generally comparable in magnitude to the peak i n fraction 6 or 7, but was markedly reduced in transition cells grown with glucose. It should be noted that none of the other hydrolase activities assayed in the present work or in previous studies by Muller (‘71) or by Rothstein and Blum (’74) ever appeared in a particle which barely moved

465

LYSOSOMAL HYDROLASE RELEASE

I

/

12,OOOr I0,ooO 0)

g c

-

cid phosphatase

I

I

I

7120

- 100

I

0

8000-

-80

S 6000-

-60

n

5

Q

0 4000-

-40

8

- 20

2000 -

m c

9

2

40

30 20

10

680

0

640

280 -

P - N - a c e t y l hexoseaminidase

FRACTION NUMBER Fig. 3 Distribution of six hydrolase activities i n sucrose gradients of homogenates from stationary phase cultures. Cells were grown for 43 hours from a n initial density of 59,000 cellslml t o a final density of 739,000 cells/ml, and then collected, homogenized, and layered onto a sucrose density gradient as described in MATERIALS AND METHODS. Units are as i n the legend to figure 2 . Percent recoveries were. acid protease, 99% ; p-fucosidase, 101 % ; P-galactosidase, 143%; a-mannosidase, 110 % , p-N-acetylhexoseaminidase, 122 5% ; acid phosphatase, 115%.

into the gradient in a 4-hour centrifugation.

p-N-acetylhexoseaminidase Hexoseaminidase activity occurred i n two locations in the gradient. One peak of activity was always localized with the main lysosomal peak at about fraction 6 or 7. A second peak of activity at about fkactions 11-13 was always present in transition cells, whether grown in the presence or

absence of glucose, but was either greatly reduced or completely absent in cells grown to the stationary phase (compare fig. 3 with figs. 2, 4). It was of interest to inquire whether isoenzymes A and B were present throughout the gradient in the same ratios as found in the secreted hexoseaminidase as analyzed by DEAE column chromatography. For this purpose, we assume that the properties of the A and B forms are as shown i n table 2, and ask

466

J . J . BLUM

5oooc a, Y)

2 4000r 0 a

6

A Acid phosphotose

I\

1

I60 a,

3000-

rI r\ - b

i

L

a 2000 D

120

g

c

80

1000-

JO

O L

6ot 50

6

P-galactosidase

t

240 -

200 160-

K-

N-acetyl hexoseaminidase

120-

\

f

F R A C T I O N NUMBER Fig. 4 Distribution of six hydrolase activities i n sucrose gradients of homogenates from cultures grown to transition phase with 15.4 mM glucose. Cells were grown for 18 hours from a n initial density of 48,200 cellslml to a final density of 670,000 cellslml and then collected, homogenized, and layered onto a sucrose density gradient as described i n MATERIALS A N D METHODS. Units are as in the legend to figure 2 . Percent recoveries were: acid protease, 132% ; p-fucosidase, 130% ; P-galactosidase, 9 3 % ; a-mannosidase, 135% ; 6-N-acetyl-hexoseaminidase, 119% ; acid phosphatase, 9 4 % .

what mixture of these two would have the properties of the hexoseaminidase in fractions 6 or 7 (heavy lysosomes), 11 or 12 (light lysosomes), and 18 or 19 (“free” hydrolase activity) as shown in table 3. Examination of the pH 6 property of the lysosomal hexoseaminidase in peaks 5,6 of the sucrose gradient indicate that it corresponds to a mixture of about 36% and 64% of hexoseaminidase A and B, respectively. Given the uncertainty in the

measurements, this is fully compatible with the proportions of these isoenzymes found in the DEAE gradients. Its heat stability property, however, is compatible with its containing only the B isoenzyme. Similarly, hexoseaminidase in the lower density lysosomal peaks (fractions 11,12 of the sucrose gradient) responds to raising the pH of assay to 6.0 as if this peak consisted of a mixture of 55% and 45% of hexoseaminidase A and B, respectively,

35.1 t 5.9 (n = 7) 90.4 5.0 (n = 8)

91.8 +- 7.4 (n = 11)

11,12

43.6 2 8.3 ( n = 11)

6,7

68.7 t 17.7 ( n = 11)

70.2 2 12.1 (n = 9)

8.8 k 3.9 (n = 9)

8.0 +- 1.8 (n = 10)

73.5 +- 16.4 (n = 11)

17.6 & 3.6 ( n = 11)

(n = 8) (n = 8)

78.8 t 0.8 ( n = 11)

41.9 5 4 . 7

44.5 t 13.6

18,19

(n=11)

11,12

56.9 C 18.3

% activity remaining

6,7

(n=11)

18,19

Acid phosphatase

59.2 t 12.3

8-N-acetylhexoseaminidase

* 4.0 (n = 3)

98.3

18,19 (18 hr+glucose)*

Samples were taken from sucrose gradients such as those shown i n figures 2, 3, and 4 from t h e tubes with maximum activity for that hydrolase i n the position shown. Samples taken from fraction 6 or 7 refer, therefore, to the heavy band of lysosomes, banding at a density of about 1.23 gmicrns, those from fraction 11 or 12 (or i n some cases 10 or 13) to the light population of lysosomes, banding a t a density of about 1.16 gmlcrna; those from fraction 18 or 19 to the “free” hydrolase on top of t h e gradient. The various fractions were then assayed a s described i n MATEnxALs A N D ~ ~ E T H O D(in S the presence of Triton X-100) at p H 4.5 ( = 100% activity for each fraction) a n d under the conditions shown i n the table. Data from the various growth conditions have been pooled except for the p H 6.0 d a t a on the “free” phosphatase from cells grown to transition phase i n the presence of glucose.

0.2 mM PCMB

10 min at 66O

Treatment p H 6.0

Fraction No.

TABLE 3

Sensitivity of acid phosphatase and ofJ3-N-acetylhexoseaminidase activities from sucrose density gradients to heat, pH, and pchloromercuribenzoate

M

m s m r

v1

%

r

s 0

tf

4

X

$

z

d

r

468

J. J. BLUM

but the thermolability property corresponds to that of isoenzyme B only. The “free” hexoseaminidase (sucrose gradient fractions 18,19) consists entirely of isoenzyme B according to the pH 6 criterion and at least 85% of isoenzyme B according to thermolability criterion. Thus the intracellular hexoseaminidase(s) nowhere resembles the secreted hexoseaminidases in its thermolability properties. Furthermore, although the intralysosomal hexoseaminidase(s) could be a mixture of the secreted isoenzymes i n the proportion found by DEAE chromatography as far as the pH 6 criterion is concerned, the “free” hexoseaminidase of the homogenate does not behave at pH 6 as would be expected for a mixture of the isoenzymes. Acid phosphatase This hydrolase is found i n both the light (peaks 11,12) and heavy b e a k s 5,s) lysosomal fraction and “free” i n the non-particulate portion of the homogenate as well (peaks 18,19). As with p-N-acetylhexoseaminidase and a-mannosidase, the activity in the light lysosomal peak is greatly reduced in cells grown to stationary phase as compared to transition phase (compare fig. 3 with figs. 2,4). Although it appears from figure 2 that there was more acid phosphatase in the light lysosomal fraction than in the heavy in transition cells, this is not generally true; usually there was less activity in the light lysosomal peak than in the heavy lysosomal peak. It should be noted that with increasing culture age the intracellular content of acid phosphatase increases (Rothstein and Blum, ’74a). Comparison of the curves for acid phosphatase activity in the homogenates of Tetrahymena (figs. 2,3) indicates that much of the increase occurs in the heavy lysosomal fraction. It is not known whether the “free” acid phosphatase occurs as such in the cells or arises from damage to the lysosomes during homogenization. Table 3 presents data on the amount of acid phosphatase actively remaining at the three locations of the sucrose gradient after heating for ten minutes at 6 ” , in the presence of 0.2 mM p-chloromercuribenzoate, or at pH 6.0, compared to the value obtained under the standard (pH 4.5) assay conditions ( = 100%). Except

for the transition cells grown with glucose, there was no difference between transition cells and stationary cells and all the data have been pooled. The “free” acid phosphatase activity (peaks 18,19) in each of the three experiments on cells grown to transition phase in the presence of glucose, however, differed significantly in its response to a change of pH, and these data have been placed in a separate column in table 3. In principle one can compare the secreted with the intracellular acid phosphatases using the same approach as used for comparison of the secreted hexoseaminidases with those present intracellularly. In practice, however, the acid phosphatase isoenzymes are too similar with respect to their inhibition by PCMB for this criterion to be of much value, and since there are three (or more) acid phosphatase isoenzymes, one cannot draw conclusions with the same confidence as was possible for the hexoseaminidases. According to table 1, about 45% of the secreted acid phosphatase was resolved on DEAE as acid phosphatase I, and 55% as a mixture of acid phosphatases I1 and 111. Examination of the DEAE chromatograms suggests that not less than 25% of the mixed peak was either I1 or 111. We have therefore computed the properties of the intracellular peaks in two ways: (a) as if they were a mixture of 45% acid phosphatase I, 14% acid phosphatase 11, and 41 % acid phosphatase 111; (b) 45% I, 41% 11, 14% 111. Computation (a) predicts that 46% of the activity of such a mixture would be assayed at pH 6 as compared to pH 4.5, while computation (b) predicts that 42% of the activity would be assayed at pH 6. The observed values for peaks 5,6, 11,12, and 18,19 (except for transition cells grown with glucose) were 56.6 & 18.3, 44.5 f 13.6, and 41.9 f 4.7, respectively. Clearly the acid phosphatase at each location in the sucrose gradient could consist of a mixture of the acid phosphatase isoenzymes in roughly the proportions secreted. In the case of transition cells grown with glucose, however, it is clear that activity of the “free” acid phosphatase in the homogenate (fractions 18,19, table 3) changes much less in response to a change in pH from 4.5 to 6.0 than that of any of

LYSOSOMAL HYDROLASE RELEASE

469

the secreted phosphatase isoenzymes Tetrahymena enzyme is a single relative(table 2). ly non-specific p-glycosidase. This appears According to the thermostability cri- somewhat unlikely, however, since the terion computation (a) predicts that 13% proportions of p-galactosidase and pof the pH 4.5 activity would remain after fucosidase activities secreted vary as a functen minutes at 66", while computation (b) tion of prior growth conditions and since predicts that 29% of the activity would the two activities have different percent remain. Since the "free" acid phosphatase recoveries from DEAE chromatography. in the homogenate retained 17.6% of its We have not studied the properties of the activity after heating, it could consist of p-galactosidase -p-fucosidase enzymes), a mixture of the phosphatase isoenzymes nor have we studied the properties of the in the proportions secreted. The heavy a-mannosidase released by these cells. This and light lysosomal fractions, however, hydrolase is of particular interest since retained only about 8 % of their original part of it appears to reside in a very light activity after heating (table 3). This is particle which is distinct from the heavy sufficiently below the value of 13% pre- and light lysosomal peaks which are chardicted on the basis of computation (a) as to acteristically observed i n sucrose density render it unlikely that the acid phospha- gradients of Tetrahymena (Miiller, '71; tase i n either of these lysosomal popula- Rothstein and Blum, '74b; this paper, figs. tions consists of a mixture of the isoen- 2-4). This particle does not appear to conzymes in the same proportions as secreted, tain any of the other lysosomal hydrolases but much more work would have to be done studied i n the present work, and obviously to be certain. deserves further study. Dickie and Liener ('62) first suggested DISCUSSION that the medium in which Tetrahymena Beck et al. ('68) showed that good reso- grew contained at least two different acid lution of liver lysosomal hydrolases could proteases. In a n initial study of some propbe achieved by a single chromatography erties of the secreted proteases (Blum, step with CM-cellulose, and observed mul- '75), we found that chymostatin, the most tiple peaks of several lysosomal hydrolases. effective of four proteolytic inhibitors The lysosomal hydrolases secreted by Tetra- tested, did not completely inhibit the sehymena in a very dilute salt solution are, of creted protease activity, and suggested course, ideally suited for resolution by a that there were at least two proteases sesingle step, since the total amount of pro- creted. Since then Levy et al. ('76) have tein released by about 0.5.109 cells is of studied the intracellular proteases of Tetrathe order of 3 mg, and this starting ma- hymena and have found that chymostatin terial is (presumably) almost entirely free is a n effective inhibitor of some of the of other proteins. Use of large capacity four or five proteases present. DEAE columns and a suitable salt gradient Following the discovery by Robinson and permitted u s to identify three or more peaks Stirling ('68) that there were two forms of acid protease and of acid phosphatase, of p-N-acetylhexoseaminidase activity in two peaks of p-N-acetylhexoseaminidase, human spleen, much work has been done and peaks of a-mannosidase, p-galac- on these isoenzymes. Okada and OBrien tosidase, and p-fucosidase. These peaks ('69) showed that there were two forms in were not all resolved from each other, of liver and brain as well. A hexoseaminidase course, but preliminary studies with a sec- C is also found in brain, but appears to ond step of chromatography on CM- be a microsomal rather than a lysosomal cellulose suggest that a rather high state of hydrolase (Braidman et al., '74) and still purity can be achieved in all cases except another form appears to be obtained from the fucosidase-galactosidase peak, which commercial preparations of serum albumin was not separated into two components by (Kanfer and Spielvogel, '73). The disthis step. Ockennan ('68) has shown that covery that Tetrahymena secretes two at least one isoenzyme of p-galactosidase forms of a p-N-acetylhexoseaminidase is in human liver and kidney appears to be therefore of some interest since it may capable also of hydrolyzing p-glucosidase serve to shed light on the evolutionary asand p-fucosidase, and it may be that the pect of hexoseaminidase function. Similar-

470

J. J. BLUM

ly, two forms of p-N-acetylhexoseaminidase are released into the medium in which Physarum grows (Kilpatrick and Stirling, ’75). The two forms from Physarum have pH optima near pH 4.5 but differ in the sharpness of the pH profile. In Tetrahymena p-N-acetylhexoseaminidase activity is present in both the heavy and light lysosoma1 fractions. The ratio of activities at pH 6.0 relative to pH 4.5 was not inconsistent with the properties to be expected on the basis of the proportions and properties of the secreted isoenzymes, but the heat stability properties were similar to that of the B form only. This implies that modification of the enzyme occurred either prior to secretion (when autophagy could alter the properties of the enzyme subsequently released), during the secretory process (where changes could occur as a result of dilution, for example), or after release from the cell. Although the secretion occurred i n the presence of chymostatin, which inhibits much of the proteolytic activity of Tetrahymena proteases (Blum, ’75; Levy et al., ’76), formation of the more thermolabile form may nevertheless be an artifact caused by proteolytic acitivity occurring after secretion. Alternatively, if the hexoseaminidase(s) of Tetrahymena are glycoproteins as, for example, is hexoseaminidase A of human tissue (Robinson and Stirling, ’68) then the increase in thermolability may be due to the activity of one or more of the secreted glycosidases. Until more is known about the substrate specificities of the secreted and intracellular forms and about their detailed molecular properties it cannot be decided whether the modification occurs prior to, during, or subsequent to secretion, nor whether it has any physiological significance. It is also possible that the assay procedure itself is in part responsible for the discrepancy between the properties of the intralysosomal hexoseaminidase and those of the secreted forms. Thus the thermolability assay for the intralysosomal hydrolase is, of necessity, conducted in the presence of relatively high concentrations of proteases and of other glycosidases, and modification of the hexoseaminidase could occur at this step. It is also interesting that the properties of the “free” intracellular hexoseaminidase differ from those expected for a 50-50 mix-

ture of the two forms obtained from DEAE column chromatography, and, furthermore, that they differ not only from the secreted isoenzymes but from the hexoseaminidase(s) found in either the heavy or light lysosomal fractions. If one assumes that the “free” hexoseaminidase activity in the homogenates represents activity released by breakage of lysosomes during the homogenization procedure, then this would suggest that even though the homogenization and sucrose gradient centrifugation steps were performed at 0’ , there was enough protease andlor glycosidase activity to modify the hexoseaminidase, but i n a different manner than the modification which occurs during incubation or after secretion. Further studies on the hexoseaminidase of Tetrahymena are in progress. Allen (‘68) and Allen and Weremicek (’70) report that many acid phosphatases could be visualized by gel electrophoresis on syngen 1 and related strains of Tetrahymena. Border et al. (’73) found there were 12 bands of acid phosphatase activity which could be resolved from phenoset D of Tetrahymena, to which strain HSM belongs, although in similar studies Nielsen and Andronis (’75) reported only four or five acid phosphatase isoenzymes. The present studies show that at least three and possibly five acid phosphatase isoenzymes are present in the mixture of hydrolases secreted by Tetrahymena. This is clearly different than the situation in the alga Ochromonas danica, where only one intracellular acid phosphatase and one extracellular acid phosphatase (which differs in several properties from the intracellular enzyme) have been reported (Patni and Aaronson, ’74). Klamer and Fennel (‘63) showed that there were five different acid phosphatases in homogenates of strain W of Tetrahymena and that intracellular phosphatase activity rose markedly with increasing culture age, especially in glucose-supplemented cultures as the glucose was utilized (also Lazarus and Scherbaum, ’68; Blum, ’75). Because of the presence of so many isoenzymes of acid phosphatase it is difficult to compare the intracellular profile with that secreted into the medium, and only one clear difference was observed; the “free” acid phosphatase of the homogenate differed in its

LYSOSOMAL HYDROLASE RELEASE

sensitivity to pH from that of any of the secreted isoenzymes resolved by DEAE chromatography. As in the case of the “free” hexoseaminidase of the homogenate, this may indicate that the acid phosphatase(s) are modified by proteolytic and/or glycosidase activity after the homogenization step. It should be noted that the light lysosoma1 fraction resolved on sucrose gradient does not have the same hydrolase complement as the heavy lysosomes. This further documents the heterogeneity of the lysosomes in Tetrahymena without, however, contributing to our understanding of the physiological significance of this heterogeneity. The present study raises many questions concerning the relationship between intralysosomal enzymes and their secreted counterparts. It is to be hoped that further studies may provide some of the answers to the questions raised here for the first time.

471

Braidman, I . , M. Carroll, N. Dance and D. Robinson 1974 Separation and properties o f h u m a n brain hexoseaminidase C. Biochem. J., 143: 295-301. Davies, M. 1975 The heterogeneity lysosomes. In: Lysosomes in Biology and Pathology. Vol. IV. J. T. Dingle and R. T. Dean, eds. North Holland Publishing Co., Amsterdam, pp. 305348. Dickie, N., and I. E. Liener 1962 A study of the proteolytic system of T e t r a h y m e n a pyriformis W. Biochim. Biophys. Acta, 64: 4 1 5 1 . Kanfer, J. N., and C. H. Spielvogel 1973 The inhibition of p-N-acetylhexoseaminidase by lactones. Biochim. Biophys. Acta, 327: 405411. Kilpatrick, D. C., and J. L. Stirling 1975 T w o forms of p-N-acetylhexoseaminidase from Phys a r u m p o l y c e p h a l u m . Biochem. Soc. Trans., 3 : 246-247. Klamer, B., and R. A. Fennel1 1963 Acid phosphatase activity during growth a n d synchronous division of T e t r a h y m e n a pyriformis W. Exptl. Cell Res., 29: 166-175. Lazarus, L. H., and 0. H. Scherbaum 1968 Activity of ribonuclease, acid phosphatase, and phosphodiesterase in T e t r a h y m e n a pyriformis during growth. J . Cell Biol., 36: 415-418. Levy, M. R., E. E. Sisskin and C. L. McConkey 1976 A protease that increases during a period of enzymic and metabolic adjustment in TetraACKNOWLEDGMENTS h y m e n a . Arch. Biochem. Biophys., 172: 6 3 4 4 4 7 . I am indebted to Ms. Carolyn Edwards Lloyd, D., R. Brightwell, S. E. Venables, G. I. Roach and G. Turner 1971 Subcellular fracfor excellent technical assistance. This tionation of T e t r a h y m e n a pyrzformis ST by work was supported by Grant 5 R01 zonal centrifugation: Changes i n activities and HD01269 from the National Institutes of distribution of enzymes during the growth cycle and on starvation. J. Gen. Microbiol., 65: 209Health. 223. Muller, M. 1970 Release of hydrolases by TetraLITERATURE CITED h y m e n a pyriformis. J. Protozool., 17: (Suppl.): 13. Allen, S. L. 1968 Genetic and epigenetic control 1971 Lysosomes i n T e t r a h y m e n a pyriof several isozymic systems i n T e t r a h y m e n a . f o r m i s I1 Intracellular distribution of several Ann. N.Y. Acad. Sci., 151: 190-207. acid hydrolases. Acta Biol. Acad. Sci. Hung., Allen, S. L., and S. L. Weremicek 1971 Inter22: 179-186. syngenic variations in the esterases and acid 1972 Secretion of acid hydrolases and phosphatases of T e t r a h y m e n a pyriformis. Bioits intracellular source i n Tetra h y m e n a pyrichem. Genetics, 5: 119-133. f o r m i s . J . Cell Biol., 52: 4 7 8 4 8 7 . Beck, C., S. Mahadevan, R. Brightwell, C. J. DilNielsen, P. J . , and P. T. Andronis 1975 Furlard and A. L. Tappel 1968 Chromatography ther electrophoretic characterization of strains of lysosomal enzymes. Arch. Biochem. Biophys., of T e t r a h y m e n a pyriformis. J . hotozool., 22: 128: 369-377. 185-187. Benson, J. R., and P. E. Hare 1975 ophthalaldeOckerman, P . A. 1968 Identity of p-glucohyde: Fluorogenic detection or primary amines sidase, P-xylosidase, and one of the p-galacin the picomole range. Comparison with flutosidase activities in h u m a n liver when asorescamine and ninhydrin. Proc. Nat. Acad. sayed with 4-methyl-umbelliferyl p-D-glycosides. Sci. (U.S.A.), 72: 6 1 9 4 2 2 . Studies in cases of Gauchers disease. Biochim. Blum, J. J . 1975 Effects of metabolites present Biophys. Acta, 165: 5 9 4 2 . during growth of T e t r a h y m e n a pyriformis on Okada, S., and J. S. O’Brien 1969 Tay-Sachs the subsequent secretion of lysosomal hydrolases. disease: Generalized absence of a beta-D-NJ. Cell. Physiol., 86: 131-142. acetylhexoseaminidase component. Science, Blum, J. J., and T. L. Rothstein 1975 Lyso165: 698-700. somes in T e t r a h y m e n a . I n : Lysosomes i n Biology Patni, N. J., and S. Aaronson 1974 Partial and Pathology. Vol. IV. J . T. Dingle and R. T. characterization of the intra- and extracellular Dean, eds. North Holland Publishing Co., Amacid phosphatase of an alga, Ochromonas dnnica. sterdam, pp. 3 3 4 5 . J. Gen. Microbiol., 83: 9-20. Borden, D., G. S . Whitt and D. L. Nanney 1973 Robinson, D., and J. Stirling 1968 N-Acetyl-pElectrophoretic characterization of classical Glucosaminidases in h u m a n spleen. Biochem. Tetrahymena pyriformis strains. J . Protozool., J., 107: 321-327. 20: 693-700.

472

J. J. BLUM

Rothstein, T. L., and J. J. Blum 1973 Effect of glucose, acetate, pyruvate, and carmine on i n tracellular content and extracellular release of three acid hydrolases. J. Cell Biol., 57: 630-641. 1974a Lysosomal physiology i n Tetrah y m e n a . 11. Effect of culture age and tempera-

ture on the extracellular release of 3 acid hydrolases. J. Protozool., 21 : 163-168. _ _ 1974b Lysosomal physiology in Tetrahymena. IV. Effect of dichloroisoproterenol on the intracellular source of released acid hydrolases. Exptl. Cell Res., 87: 168-174.