The Journal of Volume 39

Protozoology

November-December

Number 6

J . Protozool., 39(6), 1992. pp. 655-662 0 1992 by the Society of Protozoologists

An Assessment of Proteolytic Enzymes in Tetrahymena thermophila J. WILLIAM STRAUS,' GRACE MIGAKI and MARIA T. FINCH Department of Biology, Vassar College, Poughkeepsie, New York 12601

ABSTRACT. Cellular extracts of Tetrahymena therrnophila were found to contain substantial levels of proteolytic activity. Protein digestion occurred over broad ranges of pH, ionic strength, and temperature and was stimulated by treatment with thiol reductants, EDTA and sodium dodecyl sulfate. Incubation at temperatures 260" C or with high concentrations of chaotropic reagents such as 10 M urea or 6 M guanidine-HC1 caused an apparent irreversible loss of activity. Activity was also strongly diminished by increasing concentrations of divalent cations. Several peptide aldehydes, p-hydroxymercuribenzoate, and alkylating reagents such as iodoacetate, N-tosyl-L-lysine chloromethyl ketone, N-tosyl-L-phenylalanine chloromethyl ketone, N-methylmaleimide, and trans-epoxysuccinyl-Lleucylamido-(4-guanidino)-butanewere potent inhibitors of proteolytic activity. Aprotinin diminished activity by approximately 40% while benzamidine, 3,4-dichloroisocoumarin, and trypsin inhibitors from soy bean, lima bean, and chicken egg caused relatively modest inhibition of proteolytic activity. Phenylmethanesulfonyl fluoride had no apparent effect. Electrophoretic separation of proteins on SDSpolyacrylamide gels copolymerized with gelatin substrate revealed that at least eight active proteolytic enzymes were present in cell extracts ranging in apparent molecular weight from 45,000 to I IO.000. Five ofthese apparent proteases were detected in 70% ammonium sulfate precipitates. Gelatinase activity was not detectable when extracts were pretreated with iodoacetate or E-64, indicating that all ot- the enzymes observed in activity gels were sensitive to thiol alkylation. Cellular extracts of T. thermophila appeared to contain multiple forms of proteolytic enzymes which were stimulated by thiol reductants and inhibited by thiol modifying reagents. Accordingly, the proteolytic enzymes present in cell extracts appear to be predominantly cysteine proteinases. Key words. Cysteine proteinase, protease, proteinase inhibition.

P

[30] or are subjected to temperature shifts [I]. Grinde & Jonassen [22] demonstrated that T. therrnophila increase rates of protein turnover in response to starvation and that such turnover is diminished by proteinase inhibitors or by the addition of exogenous amino acids, particularly methionine [26]. The precise relationships between the proteolytic turnover of cellular components and cell differentiation or environmental adaptation in Tetrahymena are still obscure. Evidence has accumulated for the presence of multiple forms of cysteine proteinases in cellular extracts and secretions from T. pyrlformis. For instance, Levy & McConkey [30] and Blum [ 121 obtained several distinct protease fractions by ion exchange chromatography of cellular lysates from T. pyriformis. North & Walker [37] performed a n extensive examination of cellular extracts and reported kinetic, chromatographic, and electrophoretic evidence for multiple forms of cysteine proteinases. They also demonstrated that acid and neutral proteases had different subcellular distributions and that intracellular levels of acid and neutral proteases responded differently to culture age and starvation. Several laboratories have reported the purification and characterization of low molecular weight proteases from cellular extracts and secretions T. pyrlformzs. Dickie & Leiner [ 15, 161 isolated three apparent cysteine proteinases, one from cell extracts and two from conditioned growth medium, with approx' To whom correspondence should be addressed. Abbreviations used: BzArgNan, Na-benzoyl-~~-arginine-4-nitroaniimate molecular weights of 29,000, 17,000 and 10,000, respeclide; BzArgOEt, Na-benzoyl-L-arginine ethyl ester; CPI 1, calpain in- tively. Banno et al. [3] purified a small secreted cysteine proteinase hibitor I (N-acetyl-leucyl-leucyl-norleucinal); CPI 11, calpain inhibitor (M, 22,000-23.000) which exhibited acid a n d neutral p H optima I1 (N-acetyl-leucyl-leucyl-methioninal);DTT, dithiothreitol; E-64, against different substrates. Murricaine [33] purified a cysteine trans-epoxysuccinyl-~-leucylamido-(4-guanidino)-butane; EDTA, ethylenediaminetetraacetic acid, disodium salt; HEPES, N-(2-hydroxy- proteinase (M, 28,000) from lysosome enriched cell homogeethyl)piperazine-N'-(2-ethanesulfonic acid); PMSF, phenylmethanesul- nates which exhibited a neutral p H optimum. These latter two fonyl fluoride; SDS, sodium dodecyl sulfate; TLCK, N-tosyl-L-lysine proteases were found to be stimulated by thiol reductants and chloromethyl ketone; TPCK, N-tosyl-L-phenylalaninechloromethyl ke- inhibited by alkylating reagents, organomercurials, and peptide aldehydes. tone.

ROTEOLYTIC enzymes are crucial to many cellular processes, including protein turnover, post-translational modification; hormone and enzyme activation, nutrient processing, and differentiation [5, 8, 10, 13, 35, 381. In spite of the pivotal and diverse roles that proteases play in cellular physiology, there is still much to learn concerning the functions of these enzymes and the mechanisms by which cells regulate their activity, particularly in ciliate protozoans such as Tetrahymena. Tetrahymenu exhibit numerous physiological responses to environmental alterations [17, 241 which can include changes in proteolytic enzymes. Levels of intracellular and secreted proteases have been shown to vary during growth [ 2 , 34, 37, 431 and in response t o changes in the external mileau [ I , 11, 15, 22, 26, 29, 30, 32, 371. When deprived of nutrients, ciliate protozoans such as Tetrahymena respond by altering metabolism and, in micronucleate strains, by shifting from vegetative to conjugative reproduction [ 141. Starvation causes T. pyrlformis cells to increase rates of intracellular protein a n d organelle degradation [2 7-29. 3 I], and to increase secretion of hydrolytic a n d proteolytic enzymes [3 1, 39,431. Tetrahymena pyrlformis protease levels also vary when cultures are shifted from static to shaking conditions

655

656

J. PROTOZOOL., VOL. 39, NO. 6, NOVEMBER-DECEMBER 1992

Table 1. Effects of selected reagents on proteolytic activity."

suspension and recentrifugation in distilled-deionized water. Washed cells were resuspended in ice cold 50 mM sodium citrate 010 buffer, pH 3.5, containing 0.19'0 Triton X- 100 and I niM EDTA. Control Cells were then subjected to a single freeze-thaw cycle (-20" C Reagent O.D.,,,/h activity to 4" C) followed by sonication (two 5-sec cycles at 100 W at Control 0.287 k 0.010 100 4" C). Cellular homogenates were clarified by centrifugation at 1 mM dithiothreitol 0.579 0.018 202 40,000 g for 30 rnin at 4" C and dialyzed against 0.32 mM HCl 1 mM P-mercaptoethanol 0.367 f 0.01 1 128 at 4" C to at least a 106-fold dilution using cellulose dialysis 1 O/o SDS 0.509 k 0.012 177 tubing with a molecular weight cutoffof 12,000 to 14,000 (Spec0.1% Triton X-100 0.247 f 0.009 86 trum Medical Industries, Los Angeles, CA). Extracts were then Control 0.268 f 0.008 100 filtered through 0.45-pm cellulose acetate membranes and stored 10 mM benzamidine 0.247 5z 0.002 92 either at 4" C over brief periods or frozen at -20" C for storage 0.231 f 0.004 1 mM 3,4-dichloroisocoumarin 86 up to eight weeks. Partial purification of proteases from these 1 m M EDTA 0.372 0.005 139 cell extracts was performed by sequential precipitation with 1 mM p-hydroxymercuribenzoate 7 0.020 k 0.003 ammonium sulfate in 10°/o concentration intervals over a range 0.032 k 0.001 1 mM iodoacetate 12 of 0% to 100% [ 181. Precipitates were collected by centrifugation 7 1 mM E-64 0.019 & 0.003 at 40,000 g for 15 rnin at 4" C, redissolved in 50 mM sodium 1 mM N-methylmaleimide 0.086 k 0.003 32 citrate, pH 3.5, and dialyzed as above. Protein concentrations 0.265 f 0.009 1 mM PMSF 100 1 mM TLCK 0.020 f 0.002 7 were estimated spectrophotometrically at 562 nm using bicin1 m M TPCK 17 0.046 k 0.003 choninic acid (BCA) protein assay reagent (Pierce Chemical Co., Rockford, IL) against a bovine serum albumin standard [4 I]. a Aliquots (25 PI) of cell extract were preincubated with reagents for Proteolytic assays. Proteolytic activity in cellular extracts 5 rnin in 10 mM HEPES, pH 7.0, in a 0.45-ml volume. Reactions were initiatcd by addition of 0.05 ml 1% azocasein and incubated for 1 h at was measured spectrophotometrically at 440 nm using azoal35" c . bumin and azocasein as substrates [42]. Reactions were routinely performed in triplicate at 35" C in 10 mM HEPES, 1 mM EDTA, 1 mM DTT, and 0.1% substrate (w/v) unless otherwise Characterization of cellular proteases is essential for eluci- indicated, and were terminated by the addition of one volume dating the complex roles that these enzymes play in cellular of cold 10% trichloroacetic acid. After incubation on ice for 10 responses to growth and environment. Here we present evidence min, precipitates were removed by centrifugation at 15,000 g that cellular extracts of T. thermophila contain a variety of pro- for 4 min. The supernatants were neutralized by the addition teolytic enzymes which are active over broad ranges of pH, of an equal volume of 1 M NaOH and the absorbances were temperature, and ionic strength, and exhibit high sensitivity to determined. The rates of azoalbumin and azocasein hydrolysis thiol-reducing reagents, alkylating reagents, peptide aldehyde were constant over periods exceeding 2 h, when aliquots of cell inhibitors, and a variety of other reagents. Most of the proteo- extracts containing 16-23 pg protein were assayed under the above conditions. For experiments, reactions were terminated lytic activity appears to involve cysteine proteinases. after incubations of 1 h. Azocoll (5 mglreaction) was assayed MATERIALS AND METHODS under identical conditions except that the reactions were terMaterials. Tetrahymena thermophila (strain WH- 14) was minated by filtration over glass wool to remove undigested maobtained from American Type Culture Collection (Rockville, terial and the filtrate absorbances were measured at 520 nm MD). Antipain, bestatin, chymostatin, CPI I, CPI 11, elastatinal, [421. BzArgNan hydrolysis by cell extracts was monitored spectroleupeptin, pepstatin, TLCK, and TPCK were obtained from Boehringer Mannheim Biochemicals (Indianapolis, IN). Apro- photometrically at 410 nm. An extinction coefficient of 8,800 tinin, a,-proteinase inhibitor, lima bean inhibitor, soy bean in- was used to determine nitroanaline concentrations [ 191. The hibitor, ovomuciod inhibitor, ovoinhibitor, PMSF, benzami- BzArgNan was prepared fresh by first dissolving the substrate dine-HC1, iodoacetic acid, p-hydroxymercuribenzoic acid, E-64, in dimethylsulfoxide and then diluting with 50 volumes of water N-methyl maleimide, azoalbumin, azocasein, azocoll, Bz- to yield a 4-mM stock solution. Except where indicated, assays ArgNan, BzArgOEt, and gelatin (1 75 bloom) were from Sigma were performed at 35" C in 1-ml reaction volumes containing Chemical Co. (St. Louis, MO). DTT, P-mercaptoethanol, and 10 mM HEPES, 1 mM EDTA, 1 mM DTT, and 1 mM Bzelectrophoresis reagents were obtained from BioRad (Rich- ArgNan at pH 7. For experiments, reactions were terminated mond, CA). Electrophoresis protein standards were obtained after incubations of 20-60 rnin by the addition of 0.1 ml 30% from Pharmacia LKB (Piscataway, NJ). All other chemicals acetic acid and the absorbances were measured at 41 0 nm. Hywere of reagent grade and distilled deionized water was used drolytic activity against 1 mM BzArgOEt was monitored spectrophotometrically at 253 nm as previously described [42]. throughout. The effects of various factors such as pH, temperature, chaoCell culture and extraction. Stocks of T. thermophila were maintained axenically in 10 ml of 1% proteose peptone (Difco, tropic agents, ions, detergents, thiol modifying reagents, and Detroit, MI), 0.1% yeast extract (Difco), 0.2% glucose, 0.003% protease inhibitors on proteolytic activity were evaluated. Spesequestrine NaFe (Ciba-Geigy, Ardsley, NY) growth medium cific reaction conditions are described below (Fig. 1-4; Tables at 25" C and were subcultured every two to three weeks [21]. 1, 2). Inhibitors and reagents with low aqueous solubility (CPI For experiments, cells were grown on a rotary shaker at 100 I, CPI 11, chymostatin, 3,4-dichloroisocoumarin, E-64, elastarpm, 1.9 cm orbit, in 40 ml of growth medium in 300-ml side- tinal, p-hydroxymercuribenzoate, pepstatin, PMSF, TPCK) were arm flasks or in 200 ml of medium in 1-liter flasks. Cell growth dissolved in either dimethylsulfoxide or ethanol and added in was monitored turbidometrically with a Klett-Summerson col- small aliquots to reaction mixtures. Neither solvent had aporimeter and cell density was estimated with a hemocytometer. parent effects on proteolytic activity at the concentrations used Unless otherwise noted, cells were harvested at early stationary (55% v/v). Activity gel electrophoresis. To detect native proteases in cell phase (approximately 6 x lo5cells/ml) by centrifugation at 800 g for 10 min at ambient temperature and were washed twice by extracts, discontinuous SDS-polyacrylamide gels containing gel-

*

*

657

STRAUS ET AL.-PROTEOLYTIC ENZYMES IN TETRAHYMENA THERMOPHILA J

0.60-

0.60

0

v

t

0.40

0.40

d

-

0

t t

0

d

0.20

0

0.00

2

4

6

8

10

0.20

0.00 0

40

20

60

temperature ("C) Fig. 2. Effect oftemperature on proteolytic activity. Reaction mixtures containing 25 pl cell lysate (20 pg protein) in 0.45 ml of 10 mM HEPES, 1 mM EDTA, 1 mM dithiothreitol, pH 7, were preincubated at the indicated temperatures for 5 rnin prior to addition of 50 ml 1% azocasein and then incubated for 60 min. Each point represents the average of three determinations. Standard error values did not exceed k0.020 O.D. units and were typically less.

2

a

4

10

p6H Fig. 1. Effect of pH and dithiothreitol on proteolytic activity. Reaction mixtures containing 25 pl cell lysate (2 1 r g protein), 1 mM EDTA were buffered with 0.1 M citrate from pH 2 to pH 4, 0.1 M phosphate from pH 4 to pH 8.5, and 0.1 M glycine/NaOH from pH 8.5 to pH 10 in the presence (large symbols) or absence (small symbols) of 1 mM dithiothreitol. Reactions were incubated for 60 rnin at 35" C. Each point represents the average ofthree determinations. A . Digestion ofazocasein (0)and azoalbumin (A). Standard error values did not exceed fO.015 O.D. units and were typically less. B. Hydrolysis of BzArgNan. Standard error values did not exceed k0.018 rmol/min/ml.

atin were prepared by modifications of the method of Heussen & Dowdle [25] and run on a BioRad mini-electrophoresis system. Gelatin was incorporated into 7.5% acrylamide (w/v) separating gels at a final concentration of 0.5% (w/v), a concentration that provided excellent contrast between proteolytically digested and non-digested regions in the gel upon staining. Aliquots of cell extract were diluted with one volume of loading buffer (125 mM Tris/HCl, pH 6.8,4% SDS, 4% P-mercaptoethanol, 0.004°/o bromophenol blue) and immediately loaded without heat denaturation. Electrophoresis was performed at 200 V and 4" C, and terminated when the tracking dye reached the bottom of the gel (after about 30 rnin). The gels were then immediately stained for at least 2 h with a 0.1% (w/v) solution of amido black in methanol : acetic acid : water (30: 10:60) and destained in methanol : acetic acid : water (30:10:60). For documentation, gels were photographed using high contrast film (Kodalith).

RESULTS Cell lysates typically contained 0.8 4 0.2 mglml protein following centrifugation, dialysis, and filtration. Proteolytic activity against azocasein was in the range of 0.474 k 0.107 O.D.,,, units/min/mg protein (n = 14) when extracts were assayed at pH 7.0 and 35" C in the presence of 1 m M EDTA and 1 mM DTT. The level of proteolytic activity remained stable over periods of two to three weeks when preparations were stored at pH 3.5 and 4" C. When a representative extract was assayed at neutral pH against a variety of substrates, it was found to digest azoalbumin (0.453 +- 0.005 O.D.,,, units/min/mg), azocasein (0.532 k 0.005 O.D.,,, units/min/mg), azocoll (0.267 k 0.014

Table 2. Inhibition of proteolytic activity by various peptides.a Oh Peptide (0. I mg/ml)

Control Antipain a,-proteinase inhibitor Aprotinin Bestatin CPI I CPI I1 Chymostatin Elastatinal Leupeptin Lima bean inhibitor Ovoinhibitor Ovomucoid inhibitor Pepstatin A Soy bean inhibitor

O.D.,,,/h

0.548 0.044 0.549 0.334 0.522 0.028 0.011

0.043 0.362 0.018 0.461 0.443 0.482 0.546 0.477

k 0.007 k 0.001

i 0.001 k 0.001 f 0.003 f 0.002 f 0.002 k 0.002 k 0.004 f 0.001 + 0.004 f 0.007 f 0.001 t 0.006 -t 0.005

Control activity

100

8 100 61 95 5 2 8 66 3 84 81 88 I00 87

a Aliquots (25 PI) of cell extract were preincubated with inhibitors for 5 min in 10 mM HEPES, 1 rnM EDTA, 1 mM dithiothreitol, pH 7.0, in a 0.45-mI volume. Reactions were initiated by addition of 0.05 ml 1% azocasein and incubated for 1 h at 35" C.

658

J. PROTOZOOL., VOL. 39, NO. 6 , NOVEMBER-DECEMBER 1992

0.4

0.5

0.3

0.4

1

I

0

-t

d

0

0.2

d

0.3

d

0

i d

0

0.0 O.l

0.2 0.1

0

2

4

6

chaotropic agent (M)

0.0

1

0.0

0.1

0.2

0.3

Fig. 3. Effect of chaotropic agents on proteolytic activity. Reaction mixtures containing 25 1 cell lysate ( 1 6 pg protein) were incubated for 60 min at 35" C in 50 mM HEPES, 1 mM EDTA, 1 mM dithiothreitol, Fig. 4. Effect of various salts on proteolytic activity. Reaction mixtures 0.1% azocasein, pH 7.0, with the indicated concentrations of urea (0) or guanidine-HCI (0).Each point represents the average of three de- containing 25 fil cell lysate (19 pg protein) were incubated for 60 min terminations. Standard error values did not exceed +-0.016 O.D. units at 35" C in 10 mM HEPES, 1 mM dithiothreitol, 0.1% azocasein, pH 7.0, with the indicated concentrations ofNaCl m),sodium acetate (neuand were typically less. tralized with acetic acid) (A),KCI (O), CaCI, (A), or MgClz (0).The sodium and potassium mixtures also contained 1 mM EDTA. Each point represents the average of three determinations. Standard error O.D.,,o units/min/mg), and BzArgNan (0.255 k 0.004 pmol/ values did not exceed k0.013 O.D. units and were typically less. min/mg). Hydrolysis of 1 mM BzArgOEt by cell extracts was

salt (M)

not detected. Effect of pH on proteolysis. The cell extracts exhibited proteolytic activity against azocasein, azoalbumin, and BzArgNan over a broad pH range with optimal activity for the three substrates occurring around pH 6 (Fig. 1). Removal of DTT from the assay medium caused a profound decrease in activity against each of the substrates and resulted in a slight shift of the pH optimum toward neutrality. For instance, at pH 6 the activity against azocasein, azoalbumin, and BzArgNan was diminished to 55%, 89% and 85%, respectively, when DTT was removed. The pH activity curves against the protein substrates (Fig. 1A) were broader than the activity curve against BzArgNan (Fig. 1B), and they exhibited more asymmetry in the neutral to basic range. A fourth substrate, azocoll, was also digested over a broad pH range in a pattern similar to that of azocasein and azoalbumin (not shown). EfFect of temperature. Optimal proteolytic activity against azocasein was observed at temperatures between 20" and 40" C at neutral pH, as shown in Fig. 2. The activity was decreased to around 50% of peak levels when reactions were incubated at 4" C or 50" C , and no substrate digestion was observed at or above 60" C. No proteolytic activity was detected when aliquots of lysate were heated for 10 min at or above 60" C, allowed to cool to room temperature, and then assayed aganst azocasein. Repeated freeze-thaw cycles caused a moderate loss of activity. For example, the activity of an extract with an initial azocasein digestion rate of 0.589 k 0.010 O.D. units/min/mg decreased by 18% to 0.484 k 0.02 1 O.D. units/min/mg after one freezethaw cycle. Two subsequent freeze-thaw cycles resulted in about a 6Yodecrease in activity per cycle. The level ofactivity remained stable when extracts were kept at room temperature over a 4-h period at pH 3.5 but decreased by about 12% when kept over the same period at neutral pH. Effect of chaotropic agents. Increasing concentrations of chaotropic agents resulted in diminished proteolytic activity (Fig. 3). Guanidine-HCI reduced activity in a diphasic pattern,

with an approximate 35% loss in activity between concentrations of 0 M to 0.5 M and a second notabledecrease in activity between concentrations of 2 M to 4 M. Over the same concentration range, urea caused a less pronounced decrease in proteolytic activity (e.g. 35% decrease at 6 M urea). No proteolytic activity was recovered when aliquots of cell extract were incubated overnight at 4" C in the presence of 6 M guanidine-HCI or 10 M urea (in 10 mM HEPES, pH 7 with 0.64 mg/ml protein), then dialyzed against 0.32 mM HCI as above, and assayed against azocasein in the presence of 1 mM EDTA and 1 mM DTT. Effect of monovalent and divalent ions. Sodium acetate, NaCI, and KCI slightly increased activity (